[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179335#comment-13179335 ] Phabricator commented on HBASE-4218: mcorgan has commented on the revision [jira] [HBASE-4218] HFile data block encoding framework and delta encoding implementation. Trying to review this with an eye on schema changes and compactions. INLINE COMMENTS src/main/java/org/apache/hadoop/hbase/io/hfile/HFileDataBlockEncoderImpl.java:241 What about the situation where regionserver is running for a while with ENCODING_IN_MEMORY=true and block cache gets filled with encoded blocks, and then user does schema change to disable encoding altogether. Now the block cache may return an old encoded block. (Assuming online schema change doesn't invalidate all blocks for a table?) If i'm understanding that correctly, then it shouldn't be an IllegalStateException but should be handled normally. It should probably invalidate the encoded block from the block cache if possible, otherwise it will expire normally. Then it should return null so that HfileReaderV2 knows to go to the filesystem to get the block. REVISION DETAIL https://reviews.facebook.net/D447 Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179349#comment-13179349 ] Mikhail Bautin commented on HBASE-4218: --- A brief status update. I am in the process of implementing support for column family data block encoding configuration changes. Those changes are coming in the next version of the patch that I will post tomorrow. After discussing this with Kannan, our solution is: * Assign an in-cache data block encoding to every HFile reader. This in-cache encoding is determined as follows: ** If the HFile is not encoded on disk, the in-cache encoding is set to the column family's DATA_BLOCK_ENCODING. ** If the HFile is encoded on disk, the in-cache encoding is set to the HFile encoding to avoid the wasted effort of re-encoding blocks for cache. * When a non-encoded block is loaded from disk, it is encoded using the in-cache encoding and put in cache. * When an encoded block is loaded from disk, its encoding is left as is. * To reduce the complexity of data block encoding switching, we can include the in-cache encoding type in the block cache key. For example, if ENCODED_IN_CACHE_ONLY is turned on without encoding on disk, and then the encoding is turned off altogether, the cache will be populated with non-encoded blocks (since they will have completely different keys) and encoded blocks will age out from the cache. While this is suboptimal, the implementation is very simple and the common case (when the CF encoding options do not change) is not complicated with unnecessary corner cases. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5121) MajorCompaction may affect scan's correctness
[ https://issues.apache.org/jira/browse/HBASE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179390#comment-13179390 ] Zhihong Yu commented on HBASE-5121: --- Should null assignment for lastTop be after the debug log ? MajorCompaction may affect scan's correctness - Key: HBASE-5121 URL: https://issues.apache.org/jira/browse/HBASE-5121 Project: HBase Issue Type: Bug Components: regionserver Reporter: chunhui shen Attachments: hbase-5121.patch In our test, there are two families' keyvalue for one row. But we could find a infrequent problem when doing scan's next if majorCompaction happens concurrently. In the client's two continuous doing scan.next(): 1.First time, scan's next returns the result where family A is null. 2.Second time, scan's next returns the result where family B is null. The two next()'s result have the same row. If there are more families, I think the scenario will be more strange... We find the reason is that storescanner.peek() is changed after majorCompaction if there are delete type KeyValue. This change causes the PriorityQueueKeyValueScanner of RegionScanner's heap is not sure to be sorted. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179394#comment-13179394 ] Zhihong Yu commented on HBASE-5081: --- In the above scenario, would there be many duplicate log messages in master log file ? Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3939) Some crossports of Hadoop IPC fixes
[ https://issues.apache.org/jira/browse/HBASE-3939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179450#comment-13179450 ] Benoit Sigoure commented on HBASE-3939: --- On 04/Nov/11 00:07, Stack wrote: bq. So why introduce it? Just so its in place when we want to use it later? Anyone knows the answer? Some crossports of Hadoop IPC fixes --- Key: HBASE-3939 URL: https://issues.apache.org/jira/browse/HBASE-3939 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Todd Lipcon Priority: Critical Fix For: 0.92.0 Attachments: 3939-v2.txt, 3939-v3.txt, 3939-v4.txt, 3939-v5.txt, 3939-v6.txt, 3939-v7.txt, 3939-v8.txt, 3939-v9.txt, 3939.txt A few fixes from Hadoop IPC that we should probably cross-port into our copy: - HADOOP-7227: remove the protocol version check at call time - HADOOP-7146: fix a socket leak in server - HADOOP-7121: fix behavior when response serialization throws an exception - HADOOP-7346: send back nicer error response when client is using an out of date IPC version -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Issue Comment Edited] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179394#comment-13179394 ] Zhihong Yu edited comment on HBASE-5081 at 1/4/12 2:48 PM: --- In the above scenario, would there be many duplicate log split scheduled for messages in master log file ? I think the latest patch should go through verification in a real cluster. was (Author: zhi...@ebaysf.com): In the above scenario, would there be many duplicate log messages in master log file ? Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5122) Provide flexible method for loading ColumnInterpreters
Provide flexible method for loading ColumnInterpreters -- Key: HBASE-5122 URL: https://issues.apache.org/jira/browse/HBASE-5122 Project: HBase Issue Type: Task Reporter: Zhihong Yu See the discussion on user list entitled 'AggregateProtocol Help' From Royston: {noformat} I re-created my HBase table to contain Bytes.toBytes(Long) values and that fixed it. {noformat} It was not the intention when AggregateProtocol was designed that users have to change their schema to match LongColumnInterpreter. This JIRA aims to provide a flexible way for users to load their custom ColumnInterpreters into region servers. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5123) Provide more aggregate functions for Aggregations Protocol
Provide more aggregate functions for Aggregations Protocol -- Key: HBASE-5123 URL: https://issues.apache.org/jira/browse/HBASE-5123 Project: HBase Issue Type: Improvement Reporter: Zhihong Yu Royston requested the following aggregates on top of what we already have: Median, Weighted Median, Mult See discussion entitled 'AggregateProtocol Help' on user list -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3939) Some crossports of Hadoop IPC fixes
[ https://issues.apache.org/jira/browse/HBASE-3939?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179551#comment-13179551 ] stack commented on HBASE-3939: -- @BenoƮt Is this a blocker for you? Some crossports of Hadoop IPC fixes --- Key: HBASE-3939 URL: https://issues.apache.org/jira/browse/HBASE-3939 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Todd Lipcon Priority: Critical Fix For: 0.92.0 Attachments: 3939-v2.txt, 3939-v3.txt, 3939-v4.txt, 3939-v5.txt, 3939-v6.txt, 3939-v7.txt, 3939-v8.txt, 3939-v9.txt, 3939.txt A few fixes from Hadoop IPC that we should probably cross-port into our copy: - HADOOP-7227: remove the protocol version check at call time - HADOOP-7146: fix a socket leak in server - HADOOP-7121: fix behavior when response serialization throws an exception - HADOOP-7346: send back nicer error response when client is using an out of date IPC version -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4565) Maven HBase build broken on cygwin with copynativelib.sh call.
[ https://issues.apache.org/jira/browse/HBASE-4565?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179550#comment-13179550 ] Suraj Varma commented on HBASE-4565: Ah - looks like it is out of sync with 0.92 and trunk now with the hbase-security pom updates. I'll rebase and put out a new patch version. Maven HBase build broken on cygwin with copynativelib.sh call. -- Key: HBASE-4565 URL: https://issues.apache.org/jira/browse/HBASE-4565 Project: HBase Issue Type: Bug Components: build Affects Versions: 0.92.0 Environment: cygwin (on xp and win7) Reporter: Suraj Varma Assignee: Suraj Varma Labels: build, maven Fix For: 0.94.0 Attachments: HBASE-4565-0.92.patch, HBASE-4565-v2.patch, HBASE-4565-v3-0.92.patch, HBASE-4565-v3.patch, HBASE-4565.patch This is broken in both 0.92 as well as trunk pom.xml Here's a sample maven log snippet from trunk (from Mayuresh on user mailing list) [INFO] [antrun:run {execution: package}] [INFO] Executing tasks main: [mkdir] Created dir: D:\workspace\mkshirsa\hbase-trunk\target\hbase-0.93-SNAPSHOT\hbase-0.93-SNAPSHOT\lib\native\${build.platform} [exec] ls: cannot access D:workspacemkshirsahbase-trunktarget/nativelib: No such file or directory [exec] tar (child): Cannot connect to D: resolve failed [INFO] [ERROR] BUILD ERROR [INFO] [INFO] An Ant BuildException has occured: exec returned: 3328 There are two issues: 1) The ant run task below doesn't resolve the windows file separator returned by the project.build.directory - this causes the above resolve failed. !-- Using Unix cp to preserve symlinks, using script to handle wildcards -- echo file=${project.build.directory}/copynativelibs.sh if [ `ls ${project.build.directory}/nativelib | wc -l` -ne 0]; then 2) The tar argument value below also has a similar issue in that the path arg doesn't resolve right. !-- Using Unix tar to preserve symlinks -- exec executable=tar failonerror=yes dir=${project.build.directory}/${project.artifactId}-${project.version} arg value=czf/ arg value=/cygdrive/c/workspaces/hbase-0.92-svn/target/${project.artifactId}-${project.version}.tar.gz/ arg value=./ /exec In both cases, the fix would probably be to use a cross-platform way to handle the directory locations. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5123) Provide more aggregate functions for Aggregations Protocol
[ https://issues.apache.org/jira/browse/HBASE-5123?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179559#comment-13179559 ] Tom Wilcox commented on HBASE-5123: --- SumProduct is probably another useful one. Provide more aggregate functions for Aggregations Protocol -- Key: HBASE-5123 URL: https://issues.apache.org/jira/browse/HBASE-5123 Project: HBase Issue Type: Improvement Reporter: Zhihong Yu Royston requested the following aggregates on top of what we already have: Median, Weighted Median, Mult See discussion entitled 'AggregateProtocol Help' on user list -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] gaojinchao updated HBASE-5120: -- Attachment: HBASE-5120.patch Patch is attached so that i can access it at home. Not the final one and not fully tested in cluster. Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null 2012-01-04
[jira] [Updated] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-5120: -- Comment: was deleted (was: Patch is attached so that i can access it at home. Not the final one and not fully tested in cluster.) Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null
[jira] [Updated] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-5120: -- Attachment: HBASE-5120.patch Attaching the patch so that i can access it at home. First cut versiion and not fully tested. Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null
[jira] [Updated] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-5120: -- Attachment: (was: HBASE-5120.patch) Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null 2012-01-04 00:20:39,864 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition
[jira] [Issue Comment Edited] (HBASE-5121) MajorCompaction may affect scan's correctness
[ https://issues.apache.org/jira/browse/HBASE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179390#comment-13179390 ] Zhihong Yu edited comment on HBASE-5121 at 1/4/12 5:04 PM: --- {code} +this.lastTop = null; +LOG.debug(Storescanner.peek() is changed where before = ++ this.lastTop.toString() + ,and after = {code} Should null assignment for lastTop be after the debug log ? was (Author: zhi...@ebaysf.com): Should null assignment for lastTop be after the debug log ? MajorCompaction may affect scan's correctness - Key: HBASE-5121 URL: https://issues.apache.org/jira/browse/HBASE-5121 Project: HBase Issue Type: Bug Components: regionserver Reporter: chunhui shen Attachments: hbase-5121.patch In our test, there are two families' keyvalue for one row. But we could find a infrequent problem when doing scan's next if majorCompaction happens concurrently. In the client's two continuous doing scan.next(): 1.First time, scan's next returns the result where family A is null. 2.Second time, scan's next returns the result where family B is null. The two next()'s result have the same row. If there are more families, I think the scenario will be more strange... We find the reason is that storescanner.peek() is changed after majorCompaction if there are delete type KeyValue. This change causes the PriorityQueueKeyValueScanner of RegionScanner's heap is not sure to be sorted. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179650#comment-13179650 ] Lars Hofhansl commented on HBASE-4218: -- One more thought about ENCODED_IN_CACHE_ONLY (and then I'll shut up about this)... If we ever wanted to extend this in the future and allow disk only encoding, maybe a better way would be to have ENCODING and ENCODE_ON_DISK. ENCODE_ON_DISK (default false) would just be the inverse of what ENCODED_IN_CACHE_ONLY is. That way (if we felt so inclined) we can add ENCODE_IN_CACHE later and allow it to be false. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179660#comment-13179660 ] Lars Hofhansl commented on HBASE-5120: -- Hey Ram, I know you said this is not done, yet... {code} if (t instanceof NullPointerException) { removeRegionInTransition(region); ... {code} Could we instead add a null check at the relevant point and deal with it there (or maybe throw another exception)? (Dealing with this after the NPE occurred leaves a bad taste... What we introduce another bug later that also causes an NPE, that will go unnoticed for a while.) Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition;
[jira] [Issue Comment Edited] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179660#comment-13179660 ] Lars Hofhansl edited comment on HBASE-5120 at 1/4/12 5:17 PM: -- Hey Ram, I know you said this is not done, yet... {code} if (t instanceof NullPointerException) { removeRegionInTransition(region); ... {code} Could we instead add a null check at the relevant point and deal with it there (or maybe throw another exception)? (Dealing with this after the NPE occurred leaves a bad taste... What if we introduce another bug later that also causes an NPE? That would go unnoticed for a while.) was (Author: lhofhansl): Hey Ram, I know you said this is not done, yet... {code} if (t instanceof NullPointerException) { removeRegionInTransition(region); ... {code} Could we instead add a null check at the relevant point and deal with it there (or maybe throw another exception)? (Dealing with this after the NPE occurred leaves a bad taste... What we introduce another bug later that also causes an NPE, that will go unnoticed for a while.) Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the
[jira] [Commented] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179662#comment-13179662 ] Zhihong Yu commented on HBASE-5120: --- Thanks for the quick turnaround. The patch depends on NullPointerException to detect problematic region. I think we should use more protective measure as Lars commented. The patch specifies M_ZK_REGION_CLOSING for ZKAssign.deleteNode(). This narrows the applicability of the NPE handling. Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG
[jira] [Updated] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Yu updated HBASE-5120: -- Fix Version/s: 0.94.0 0.92.0 Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Fix For: 0.92.0, 0.94.0 Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04 00:20:39,623 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null 2012-01-04 00:20:39,864 INFO org.apache.hadoop.hbase.master.AssignmentManager:
[jira] [Commented] (HBASE-5121) MajorCompaction may affect scan's correctness
[ https://issues.apache.org/jira/browse/HBASE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179666#comment-13179666 ] Lars Hofhansl commented on HBASE-5121: -- I see the patch is against trunk. Does this happen in trunk only? It might have to do with the various optimization for scanning that the FB folks put in (if I recall FB rarely does major compactions). Maybe Mikhail could have a look at this too. MajorCompaction may affect scan's correctness - Key: HBASE-5121 URL: https://issues.apache.org/jira/browse/HBASE-5121 Project: HBase Issue Type: Bug Components: regionserver Reporter: chunhui shen Attachments: hbase-5121.patch In our test, there are two families' keyvalue for one row. But we could find a infrequent problem when doing scan's next if majorCompaction happens concurrently. In the client's two continuous doing scan.next(): 1.First time, scan's next returns the result where family A is null. 2.Second time, scan's next returns the result where family B is null. The two next()'s result have the same row. If there are more families, I think the scenario will be more strange... We find the reason is that storescanner.peek() is changed after majorCompaction if there are delete type KeyValue. This change causes the PriorityQueueKeyValueScanner of RegionScanner's heap is not sure to be sorted. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179685#comment-13179685 ] Lars Hofhansl commented on HBASE-2947: -- I ran the four failing tests locally and they all pass. I would like to commit this today (the change it pretty uncontentious I think). MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Issue Comment Edited] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179685#comment-13179685 ] Lars Hofhansl edited comment on HBASE-2947 at 1/4/12 5:53 PM: -- I ran the four failing tests locally and they all pass. I would like to commit this today (the change itself is pretty uncontentious I think). was (Author: lhofhansl): I ran the four failing tests locally and they all pass. I would like to commit this today (the change it pretty uncontentious I think). MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3373) Allow regions to be load-balanced by table
[ https://issues.apache.org/jira/browse/HBASE-3373?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179712#comment-13179712 ] Zhihong Yu commented on HBASE-3373: --- @Ben, @Jonathan Gray: What do you think of my patch ? Thanks Allow regions to be load-balanced by table -- Key: HBASE-3373 URL: https://issues.apache.org/jira/browse/HBASE-3373 Project: HBase Issue Type: Improvement Components: master Affects Versions: 0.20.6 Reporter: Ted Yu Fix For: 0.94.0 Attachments: 3373.txt, HbaseBalancerTest2.java From our experience, cluster can be well balanced and yet, one table's regions may be badly concentrated on few region servers. For example, one table has 839 regions (380 regions at time of table creation) out of which 202 are on one server. It would be desirable for load balancer to distribute regions for specified tables evenly across the cluster. Each of such tables has number of regions many times the cluster size. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5097) RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE
[ https://issues.apache.org/jira/browse/HBASE-5097?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-5097: -- Status: Open (was: Patch Available) RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE --- Key: HBASE-5097 URL: https://issues.apache.org/jira/browse/HBASE-5097 Project: HBase Issue Type: Bug Components: coprocessors Reporter: ramkrishna.s.vasudevan Assignee: ramkrishna.s.vasudevan Attachments: HBASE-5097.patch, HBASE-5097_1.patch In HRegionServer.java openScanner() {code} r.prepareScanner(scan); RegionScanner s = null; if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().preScannerOpen(scan); } if (s == null) { s = r.getScanner(scan); } if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().postScannerOpen(scan, s); } {code} If we dont have implemention for postScannerOpen the RegionScanner is null and so throwing nullpointer {code} java.lang.NullPointerException at java.util.concurrent.ConcurrentHashMap.put(ConcurrentHashMap.java:881) at org.apache.hadoop.hbase.regionserver.HRegionServer.addScanner(HRegionServer.java:2282) at org.apache.hadoop.hbase.regionserver.HRegionServer.openScanner(HRegionServer.java:2272) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at org.apache.hadoop.hbase.ipc.WritableRpcEngine$Server.call(WritableRpcEngine.java:364) at org.apache.hadoop.hbase.ipc.HBaseServer$Handler.run(HBaseServer.java:1326) {code} Making this defect as blocker.. Pls feel free to change the priority if am wrong. Also correct me if my way of trying out coprocessors without implementing postScannerOpen is wrong. Am just a learner. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5097) RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE
[ https://issues.apache.org/jira/browse/HBASE-5097?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-5097: -- Attachment: HBASE-5097_2.patch Indention problem occured by mistake. Corrected patch uploaded. RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE --- Key: HBASE-5097 URL: https://issues.apache.org/jira/browse/HBASE-5097 Project: HBase Issue Type: Bug Components: coprocessors Reporter: ramkrishna.s.vasudevan Assignee: ramkrishna.s.vasudevan Attachments: HBASE-5097.patch, HBASE-5097_1.patch, HBASE-5097_2.patch In HRegionServer.java openScanner() {code} r.prepareScanner(scan); RegionScanner s = null; if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().preScannerOpen(scan); } if (s == null) { s = r.getScanner(scan); } if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().postScannerOpen(scan, s); } {code} If we dont have implemention for postScannerOpen the RegionScanner is null and so throwing nullpointer {code} java.lang.NullPointerException at java.util.concurrent.ConcurrentHashMap.put(ConcurrentHashMap.java:881) at org.apache.hadoop.hbase.regionserver.HRegionServer.addScanner(HRegionServer.java:2282) at org.apache.hadoop.hbase.regionserver.HRegionServer.openScanner(HRegionServer.java:2272) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at org.apache.hadoop.hbase.ipc.WritableRpcEngine$Server.call(WritableRpcEngine.java:364) at org.apache.hadoop.hbase.ipc.HBaseServer$Handler.run(HBaseServer.java:1326) {code} Making this defect as blocker.. Pls feel free to change the priority if am wrong. Also correct me if my way of trying out coprocessors without implementing postScannerOpen is wrong. Am just a learner. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179725#comment-13179725 ] ramkrishna.s.vasudevan commented on HBASE-5120: --- @Lars and @Ted Thanks for your reviews. I was also not convinced in adding the instanceof check for NPE. It does not look good. I was trying at other option of not calling sendRegionClose if the server is null. Also remove the RIT if anything has been added. But need to verify some more scenarios. Not sure if i can dedicate time tomorrow as we have some internal workshop. I will work on this and give my patch sooner. :) Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Fix For: 0.92.0, 0.94.0 Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in
[jira] [Commented] (HBASE-5097) RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE
[ https://issues.apache.org/jira/browse/HBASE-5097?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179727#comment-13179727 ] Andrew Purtell commented on HBASE-5097: --- +1 on fixed patch. RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE --- Key: HBASE-5097 URL: https://issues.apache.org/jira/browse/HBASE-5097 Project: HBase Issue Type: Bug Components: coprocessors Reporter: ramkrishna.s.vasudevan Assignee: ramkrishna.s.vasudevan Attachments: HBASE-5097.patch, HBASE-5097_1.patch, HBASE-5097_2.patch In HRegionServer.java openScanner() {code} r.prepareScanner(scan); RegionScanner s = null; if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().preScannerOpen(scan); } if (s == null) { s = r.getScanner(scan); } if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().postScannerOpen(scan, s); } {code} If we dont have implemention for postScannerOpen the RegionScanner is null and so throwing nullpointer {code} java.lang.NullPointerException at java.util.concurrent.ConcurrentHashMap.put(ConcurrentHashMap.java:881) at org.apache.hadoop.hbase.regionserver.HRegionServer.addScanner(HRegionServer.java:2282) at org.apache.hadoop.hbase.regionserver.HRegionServer.openScanner(HRegionServer.java:2272) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at org.apache.hadoop.hbase.ipc.WritableRpcEngine$Server.call(WritableRpcEngine.java:364) at org.apache.hadoop.hbase.ipc.HBaseServer$Handler.run(HBaseServer.java:1326) {code} Making this defect as blocker.. Pls feel free to change the priority if am wrong. Also correct me if my way of trying out coprocessors without implementing postScannerOpen is wrong. Am just a learner. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5041) Major compaction on non existing table does not throw error
[ https://issues.apache.org/jira/browse/HBASE-5041?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Shrijeet Paliwal updated HBASE-5041: Attachment: 0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch Uploading patch for 0.90 branch. Sorry for the delay Ted. Noticed your comment late that you wanted to merge it yesterday itself. Major compaction on non existing table does not throw error Key: HBASE-5041 URL: https://issues.apache.org/jira/browse/HBASE-5041 Project: HBase Issue Type: Bug Components: regionserver, shell Affects Versions: 0.90.3 Reporter: Shrijeet Paliwal Assignee: Shrijeet Paliwal Fix For: 0.92.0, 0.94.0, 0.90.6 Attachments: 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch Following will not complain even if fubar does not exist {code} echo major_compact 'fubar' | $HBASE_HOME/bin/hbase shell {code} The downside for this defect is that major compaction may be skipped due to a typo by Ops. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5124) Backport LoadTestTool to 0.92
Backport LoadTestTool to 0.92 - Key: HBASE-5124 URL: https://issues.apache.org/jira/browse/HBASE-5124 Project: HBase Issue Type: Task Reporter: Zhihong Yu Fix For: 0.92.0 LoadTestTool is very useful. This JIRA backports LoadTestTool to 0.92 so that users don't have to build TRUNK in order to use it against 0.92 cluster. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5041) Major compaction on non existing table does not throw error
[ https://issues.apache.org/jira/browse/HBASE-5041?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179735#comment-13179735 ] Hadoop QA commented on HBASE-5041: -- -1 overall. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12509436/0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch against trunk revision . +1 @author. The patch does not contain any @author tags. +1 tests included. The patch appears to include 4 new or modified tests. -1 patch. The patch command could not apply the patch. Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/664//console This message is automatically generated. Major compaction on non existing table does not throw error Key: HBASE-5041 URL: https://issues.apache.org/jira/browse/HBASE-5041 Project: HBase Issue Type: Bug Components: regionserver, shell Affects Versions: 0.90.3 Reporter: Shrijeet Paliwal Assignee: Shrijeet Paliwal Fix For: 0.92.0, 0.94.0, 0.90.6 Attachments: 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch Following will not complain even if fubar does not exist {code} echo major_compact 'fubar' | $HBASE_HOME/bin/hbase shell {code} The downside for this defect is that major compaction may be skipped due to a typo by Ops. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5094) The META can hold an entry for a region with a different server name from the one actually in the AssignmentManager thus making the region inaccessible.
[ https://issues.apache.org/jira/browse/HBASE-5094?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179738#comment-13179738 ] ramkrishna.s.vasudevan commented on HBASE-5094: --- @Stack I tried the option of applying the fix on the master side. But i felt like the history of the region was not available. (May be am wrong.) Also the ServerShutdownhandler.processShutDown() removes the online server first and also the list of regions. Then it is after the region opening is done by the balancer flow the new server name is again added back. {code} if (isServerOnline(sn)) { this.regions.put(regionInfo, sn); addToServers(sn, regionInfo); this.regions.notifyAll(); {code} bq.Could we make it such that only one thread can transition a region at a time? This i did not think much on this. @Ming Ma If it is ok, do you mind sharing the patch that you had prepared ? The META can hold an entry for a region with a different server name from the one actually in the AssignmentManager thus making the region inaccessible. Key: HBASE-5094 URL: https://issues.apache.org/jira/browse/HBASE-5094 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: ramkrishna.s.vasudevan Assignee: ramkrishna.s.vasudevan Priority: Critical Attachments: HBASE-5094_1.patch {code} RegionState rit = this.services.getAssignmentManager().isRegionInTransition(e.getKey()); ServerName addressFromAM = this.services.getAssignmentManager() .getRegionServerOfRegion(e.getKey()); if (rit != null !rit.isClosing() !rit.isPendingClose()) { // Skip regions that were in transition unless CLOSING or // PENDING_CLOSE LOG.info(Skip assigning region + rit.toString()); } else if (addressFromAM != null !addressFromAM.equals(this.serverName)) { LOG.debug(Skip assigning region + e.getKey().getRegionNameAsString() + because it has been opened in + addressFromAM.getServerName()); } {code} In ServerShutDownHandler we try to get the address in the AM. This address is initially null because it is not yet updated after the region was opened .i.e. the CAll back after node deletion is not yet done in the master side. But removal from RIT is completed on the master side. So this will trigger a new assignment. So there is a small window between the online region is actually added in to the online list and the ServerShutdownHandler where we check the existing address in AM. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5121) MajorCompaction may affect scan's correctness
[ https://issues.apache.org/jira/browse/HBASE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179743#comment-13179743 ] Todd Lipcon commented on HBASE-5121: No test for this? Can we get a functional test that can at least be run from the command line to trigger it? Nits: the javadoc on the constructors are unnecessary (they don't give any extra information) Using the exception for control flow here looks dirty IMO. MajorCompaction may affect scan's correctness - Key: HBASE-5121 URL: https://issues.apache.org/jira/browse/HBASE-5121 Project: HBase Issue Type: Bug Components: regionserver Reporter: chunhui shen Attachments: hbase-5121.patch In our test, there are two families' keyvalue for one row. But we could find a infrequent problem when doing scan's next if majorCompaction happens concurrently. In the client's two continuous doing scan.next(): 1.First time, scan's next returns the result where family A is null. 2.Second time, scan's next returns the result where family B is null. The two next()'s result have the same row. If there are more families, I think the scenario will be more strange... We find the reason is that storescanner.peek() is changed after majorCompaction if there are delete type KeyValue. This change causes the PriorityQueueKeyValueScanner of RegionScanner's heap is not sure to be sorted. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Prakash Khemani updated HBASE-5081: --- Attachment: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch implement feedback Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (HBASE-5121) MajorCompaction may affect scan's correctness
[ https://issues.apache.org/jira/browse/HBASE-5121?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan reassigned HBASE-5121: - Assignee: chunhui shen MajorCompaction may affect scan's correctness - Key: HBASE-5121 URL: https://issues.apache.org/jira/browse/HBASE-5121 Project: HBase Issue Type: Bug Components: regionserver Reporter: chunhui shen Assignee: chunhui shen Attachments: hbase-5121.patch In our test, there are two families' keyvalue for one row. But we could find a infrequent problem when doing scan's next if majorCompaction happens concurrently. In the client's two continuous doing scan.next(): 1.First time, scan's next returns the result where family A is null. 2.Second time, scan's next returns the result where family B is null. The two next()'s result have the same row. If there are more families, I think the scenario will be more strange... We find the reason is that storescanner.peek() is changed after majorCompaction if there are delete type KeyValue. This change causes the PriorityQueueKeyValueScanner of RegionScanner's heap is not sure to be sorted. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-4773) HBaseAdmin may leak ZooKeeper connections
[ https://issues.apache.org/jira/browse/HBASE-4773?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] ramkrishna.s.vasudevan updated HBASE-4773: -- Resolution: Fixed Fix Version/s: 0.92.0 Status: Resolved (was: Patch Available) Committed sometime back. HBaseAdmin may leak ZooKeeper connections - Key: HBASE-4773 URL: https://issues.apache.org/jira/browse/HBASE-4773 Project: HBase Issue Type: Bug Components: client Affects Versions: 0.90.4 Reporter: gaojinchao Assignee: xufeng Priority: Critical Fix For: 0.92.0, 0.90.6 Attachments: 4773.patch, branches_4773.patch, trunk_4773_patch.patch When master crashs, HBaseAdmin will leaks ZooKeeper connections I think we should close the zk connetion when throw MasterNotRunningException public HBaseAdmin(Configuration c) throws MasterNotRunningException, ZooKeeperConnectionException { this.conf = HBaseConfiguration.create(c); this.connection = HConnectionManager.getConnection(this.conf); this.pause = this.conf.getLong(hbase.client.pause, 1000); this.numRetries = this.conf.getInt(hbase.client.retries.number, 10); this.retryLongerMultiplier = this.conf.getInt(hbase.client.retries.longer.multiplier, 10); //we should add this code and close the zk connection try{ this.connection.getMaster(); }catch(MasterNotRunningException e){ HConnectionManager.deleteConnection(conf, false); throw e; } } -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179763#comment-13179763 ] Kannan Muthukkaruppan commented on HBASE-4218: -- I also like ENCODE_ON_DISK instead of ENCODE_IN_CACHE_ONLY (with the reverse semantics). I would say let's keep the default for ENCODE_ON_DISK to true though. This is more a testing knob in early stages-- where someone will set it to false before publishing a new data block encoder for general use. By the time end users try this, the code should be robust enough, and the Column Family setting of which data block encoding to use should be ideally the only knob they need to think about. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-3373) Allow regions to be load-balanced by table
[ https://issues.apache.org/jira/browse/HBASE-3373?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179767#comment-13179767 ] Lars Hofhansl commented on HBASE-3373: -- I forgot to +1 it above. Allow regions to be load-balanced by table -- Key: HBASE-3373 URL: https://issues.apache.org/jira/browse/HBASE-3373 Project: HBase Issue Type: Improvement Components: master Affects Versions: 0.20.6 Reporter: Ted Yu Fix For: 0.94.0 Attachments: 3373.txt, HbaseBalancerTest2.java From our experience, cluster can be well balanced and yet, one table's regions may be badly concentrated on few region servers. For example, one table has 839 regions (380 regions at time of table creation) out of which 202 are on one server. It would be desirable for load balancer to distribute regions for specified tables evenly across the cluster. Each of such tables has number of regions many times the cluster size. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179789#comment-13179789 ] Hadoop QA commented on HBASE-5081: -- -1 overall. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12509441/0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch against trunk revision . +1 @author. The patch does not contain any @author tags. +1 tests included. The patch appears to include 8 new or modified tests. -1 javadoc. The javadoc tool appears to have generated -151 warning messages. +1 javac. The applied patch does not increase the total number of javac compiler warnings. -1 findbugs. The patch appears to introduce 77 new Findbugs (version 1.3.9) warnings. +1 release audit. The applied patch does not increase the total number of release audit warnings. -1 core tests. The patch failed these unit tests: org.apache.hadoop.hbase.mapreduce.TestImportTsv org.apache.hadoop.hbase.mapred.TestTableMapReduce org.apache.hadoop.hbase.mapreduce.TestHFileOutputFormat Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/665//testReport/ Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/665//artifact/trunk/patchprocess/newPatchFindbugsWarnings.html Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/665//console This message is automatically generated. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5109) Fix TestAvroServer so that it waits properly for the modifyTable operation to complete
[ https://issues.apache.org/jira/browse/HBASE-5109?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179794#comment-13179794 ] Ming Ma commented on HBASE-5109: Thanks, Ted. Then, it is mostly taken cared of. Just want to point out two small issues in the fix for HBase-4621: 1. The loop won't stop if there is a bug in modifyTable call. In that case, the unit test will just hang instead of fail. Adding a timeout might be helpful. 2. The two asserts statement after the while statement are redundant. So we can take one out. Fix TestAvroServer so that it waits properly for the modifyTable operation to complete -- Key: HBASE-5109 URL: https://issues.apache.org/jira/browse/HBASE-5109 Project: HBase Issue Type: Bug Components: test Reporter: Ming Ma Assignee: Ming Ma Attachments: HBASE-5109-0.92.patch TestAvroServer has the following issue impl.modifyTable(tableAname, tableA); // It can take a while for the change to take effect. Wait here a while. while(impl.describeTable(tableAname) == null ) { Threads.sleep(100); } assertTrue(impl.describeTable(tableAname).maxFileSize == 123456L); impl.describeTable(tableAname) returns the default maxSize 256M right away as modifyTable is async. Before HBASE-4328 is fixed, we can fix the test code to wait for say max of 5 seconds to check if impl.describeTable(tableAname).maxFileSize is uploaded to 123456L. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179797#comment-13179797 ] stack commented on HBASE-5081: -- These failures seem unrelated to this patch -- they happened on previous hadoopqa build for unrelated patch. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179798#comment-13179798 ] stack commented on HBASE-5081: -- I'll commit this unless objection. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179802#comment-13179802 ] Matt Corgan commented on HBASE-4218: Some food for thought - there is probably more complexity to this down the road. There are always going to be trade-offs between encoding speed, compression ratio, scan throughput, and seek latency. These trade-offs can actually be quite huge, like 10x when you start considering things like suffix compression. I can see having different encodings in the same column family depending on dynamic performance decisions. For example, use the most compact encoding during major compaction, but use the fastest encoding if memstore flushes are backlogged. We probably can't get it perfect in this first iteration. Just want to avoid shooting ourselves in the foot as much as possible. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5109) Fix TestAvroServer so that it waits properly for the modifyTable operation to complete
[ https://issues.apache.org/jira/browse/HBASE-5109?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179805#comment-13179805 ] stack commented on HBASE-5109: -- +1 on committing the patch. Its some nice, if small, cleanup. Fix TestAvroServer so that it waits properly for the modifyTable operation to complete -- Key: HBASE-5109 URL: https://issues.apache.org/jira/browse/HBASE-5109 Project: HBase Issue Type: Bug Components: test Reporter: Ming Ma Assignee: Ming Ma Attachments: HBASE-5109-0.92.patch TestAvroServer has the following issue impl.modifyTable(tableAname, tableA); // It can take a while for the change to take effect. Wait here a while. while(impl.describeTable(tableAname) == null ) { Threads.sleep(100); } assertTrue(impl.describeTable(tableAname).maxFileSize == 123456L); impl.describeTable(tableAname) returns the default maxSize 256M right away as modifyTable is async. Before HBASE-4328 is fixed, we can fix the test code to wait for say max of 5 seconds to check if impl.describeTable(tableAname).maxFileSize is uploaded to 123456L. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179806#comment-13179806 ] Zhihong Yu commented on HBASE-5081: --- {code} +if (oldtask.status == FAILURE) { + // wait for status to change to DELETED + try { +oldtask.wait(); + } catch (InterruptedException e) { +LOG.warn(Interrupted when waiting for znode delete callback); +// fall through to return failure {code} Should a loop be introduced wrapping oldtask.wait() ? JVM sometimes produces spurious notification. {code} +LOG.fatal(Logic error. Deleted task still present in tasks map); +assert false : Deleted task still present in tasks map; {code} The assertion wouldn't be effective at runtime, right ? I think throwing exception would be better. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5093) wiki update for HBase/Scala
[ https://issues.apache.org/jira/browse/HBASE-5093?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179809#comment-13179809 ] stack commented on HBASE-5093: -- I just added you as a wiki contributor Joe. Let me know if you still can't edit. wiki update for HBase/Scala --- Key: HBASE-5093 URL: https://issues.apache.org/jira/browse/HBASE-5093 Project: HBase Issue Type: Improvement Reporter: Joe Stein I tried to edit the wiki but it says immutable page would be helpful/nice for folks to know how to get sbt working with Scala the following is what I did to get it working, not sure why could not edit the wiki figure i open a JIRA so someone with access could update this {code} resolvers += Apache HBase at https://repository.apache.org/content/repositories/releases; resolvers += Thrift at http://people.apache.org/~rawson/repo/; libraryDependencies ++= Seq( org.apache.hadoop % hadoop-core % 0.20.2, org.apache.hbase % hbase % 0.90.4 ) {code} or let me access it and I can do it, np -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179811#comment-13179811 ] Zhihong Yu commented on HBASE-5081: --- See http://stackoverflow.com/questions/1050592/do-spurious-wakeups-actually-happen Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179813#comment-13179813 ] stack commented on HBASE-4218: -- /me hearts this issue Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179821#comment-13179821 ] Zhihong Yu commented on HBASE-5081: --- w.r.t. master retry after the following exception is raised: {code} throw new IOException(duplicate log split scheduled for + lf.getPath()); {code} I see this code in MasterFileSystem.splitLog(): {code} try { splitLogSize = splitLogManager.splitLogDistributed(logDirs); } catch (OrphanHLogAfterSplitException e) { LOG.warn(Retrying distributed splitting for + serverNames, e); splitLogManager.splitLogDistributed(logDirs); } {code} Would the retry be carried out as described ? Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5118) Fix Scan documentation
[ https://issues.apache.org/jira/browse/HBASE-5118?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179826#comment-13179826 ] stack commented on HBASE-5118: -- +1 Fix Scan documentation -- Key: HBASE-5118 URL: https://issues.apache.org/jira/browse/HBASE-5118 Project: HBase Issue Type: Sub-task Components: documentation Affects Versions: 0.94.0 Reporter: Lars Hofhansl Assignee: Lars Hofhansl Priority: Trivial Attachments: 5118.txt Current documentation for scan states: {code} Scan scan = new Scan(); scan.addColumn(Bytes.toBytes(cf),Bytes.toBytes(attr)); scan.setStartRow( Bytes.toBytes(row)); // start key is inclusive scan.setStopRow( Bytes.toBytes(row + new byte[] {0})); // stop key is exclusive for(Result result : htable.getScanner(scan)) { // process Result instance } {code} row + new byte[] {0} is not correct. That should row + (char)0 -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179825#comment-13179825 ] stack commented on HBASE-2947: -- Is this safe to do Lars? {code} -public class Increment implements Writable { +public class Increment implements Row { {code} Should you leave Writable in place? Else LGTM MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5097) RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE
[ https://issues.apache.org/jira/browse/HBASE-5097?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179834#comment-13179834 ] stack commented on HBASE-5097: -- +1 on patch RegionObserver implementation whose preScannerOpen and postScannerOpen Impl return null can stall the system initialization through NPE --- Key: HBASE-5097 URL: https://issues.apache.org/jira/browse/HBASE-5097 Project: HBase Issue Type: Bug Components: coprocessors Reporter: ramkrishna.s.vasudevan Assignee: ramkrishna.s.vasudevan Attachments: HBASE-5097.patch, HBASE-5097_1.patch, HBASE-5097_2.patch In HRegionServer.java openScanner() {code} r.prepareScanner(scan); RegionScanner s = null; if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().preScannerOpen(scan); } if (s == null) { s = r.getScanner(scan); } if (r.getCoprocessorHost() != null) { s = r.getCoprocessorHost().postScannerOpen(scan, s); } {code} If we dont have implemention for postScannerOpen the RegionScanner is null and so throwing nullpointer {code} java.lang.NullPointerException at java.util.concurrent.ConcurrentHashMap.put(ConcurrentHashMap.java:881) at org.apache.hadoop.hbase.regionserver.HRegionServer.addScanner(HRegionServer.java:2282) at org.apache.hadoop.hbase.regionserver.HRegionServer.openScanner(HRegionServer.java:2272) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at org.apache.hadoop.hbase.ipc.WritableRpcEngine$Server.call(WritableRpcEngine.java:364) at org.apache.hadoop.hbase.ipc.HBaseServer$Handler.run(HBaseServer.java:1326) {code} Making this defect as blocker.. Pls feel free to change the priority if am wrong. Also correct me if my way of trying out coprocessors without implementing postScannerOpen is wrong. Am just a learner. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5071) HFile has a possible cast issue.
[ https://issues.apache.org/jira/browse/HBASE-5071?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179838#comment-13179838 ] stack commented on HBASE-5071: -- lgtm HFile has a possible cast issue. Key: HBASE-5071 URL: https://issues.apache.org/jira/browse/HBASE-5071 Project: HBase Issue Type: Bug Components: io Affects Versions: 0.90.0 Reporter: Harsh J Labels: hfile HBASE-3040 introduced this line originally in HFile.Reader#loadFileInfo(...): {code} int allIndexSize = (int)(this.fileSize - this.trailer.dataIndexOffset - FixedFileTrailer.trailerSize()); {code} Which on trunk today, for HFile v1 is: {code} int sizeToLoadOnOpen = (int) (fileSize - trailer.getLoadOnOpenDataOffset() - trailer.getTrailerSize()); {code} This computed (and casted) integer is then used to build an array of the same size. But if fileSize is very large ( Integer.MAX_VALUE), then there's an easy chance this can go negative at some point and spew out exceptions such as: {code} java.lang.NegativeArraySizeException at org.apache.hadoop.hbase.io.hfile.HFile$Reader.readAllIndex(HFile.java:805) at org.apache.hadoop.hbase.io.hfile.HFile$Reader.loadFileInfo(HFile.java:832) at org.apache.hadoop.hbase.regionserver.StoreFile$Reader.loadFileInfo(StoreFile.java:1003) at org.apache.hadoop.hbase.regionserver.StoreFile.open(StoreFile.java:382) at org.apache.hadoop.hbase.regionserver.StoreFile.createReader(StoreFile.java:438) at org.apache.hadoop.hbase.regionserver.Store.loadStoreFiles(Store.java:267) at org.apache.hadoop.hbase.regionserver.Store.init(Store.java:209) at org.apache.hadoop.hbase.regionserver.HRegion.instantiateHStore(HRegion.java:2088) at org.apache.hadoop.hbase.regionserver.HRegion.initialize(HRegion.java:358) at org.apache.hadoop.hbase.regionserver.HRegion.openHRegion(HRegion.java:2661) at org.apache.hadoop.hbase.regionserver.HRegion.openHRegion(HRegion.java:2647) {code} Did we accidentally limit single region sizes this way? (Unsure about HFile v2's structure so far, so do not know if v2 has the same issue.) -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179841#comment-13179841 ] Zhihong Yu commented on HBASE-5081: --- About the test suite, from https://builds.apache.org/job/PreCommit-HBASE-Build/665/console: {code} Running org.apache.hadoop.hbase.master.TestSplitLogManager Running org.apache.hadoop.hbase.regionserver.TestSeekOptimizations Tests run: 6, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 3.055 sec {code} It is a pity that the hung test was masked by known failures. MAPREDUCE-3583 seems to have gone into oblivion. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5088) A concurrency issue on SoftValueSortedMap
[ https://issues.apache.org/jira/browse/HBASE-5088?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179843#comment-13179843 ] stack commented on HBASE-5088: -- +1 on last patch posted by Lars. A concurrency issue on SoftValueSortedMap - Key: HBASE-5088 URL: https://issues.apache.org/jira/browse/HBASE-5088 Project: HBase Issue Type: Bug Components: client Affects Versions: 0.90.4, 0.94.0 Reporter: Jieshan Bean Assignee: Jieshan Bean Attachments: 5088-useMapInterfaces.txt, 5088.generics.txt, HBase-5088-90.patch, HBase-5088-trunk.patch, HBase5088Reproduce.java SoftValueSortedMap is backed by a TreeMap. All the methods in this class are synchronized. If we use this method to add/delete elements, it's ok. But in HConnectionManager#getCachedLocation, it use headMap to get a view from SoftValueSortedMap#internalMap. Once we operate on this view map(like add/delete) in other threads, a concurrency issue may occur. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4397) -ROOT-, .META. tables stay offline for too long in recovery phase after all RSs are shutdown at the same time
[ https://issues.apache.org/jira/browse/HBASE-4397?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179846#comment-13179846 ] stack commented on HBASE-4397: -- Nice one Ming. -ROOT-, .META. tables stay offline for too long in recovery phase after all RSs are shutdown at the same time - Key: HBASE-4397 URL: https://issues.apache.org/jira/browse/HBASE-4397 Project: HBase Issue Type: Bug Reporter: Ming Ma Assignee: Ming Ma Fix For: 0.92.0, 0.94.0 Attachments: HBASE-4397-0.92.patch 1. Shutdown all RSs. 2. Bring all RS back online. The -ROOT-, .META. stay in offline state until timeout monitor force assignment 30 minutes later. That is because HMaster can't find a RS to assign the tables to in assign operation. 011-09-13 13:25:52,743 WARN org.apache.hadoop.hbase.master.AssignmentManager: Failed assignment of -ROOT-,,0.70236052 to sea-lab-4,60020,1315870341387, trying to assign elsewhere instead; retry=0 java.net.ConnectException: Connection refused at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method) at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:567) at org.apache.hadoop.net.SocketIOWithTimeout.connect(SocketIOWithTimeout.java:206) at org.apache.hadoop.net.NetUtils.connect(NetUtils.java:373) at org.apache.hadoop.hbase.ipc.HBaseClient$Connection.setupIOstreams(HBaseClient.java:345) at org.apache.hadoop.hbase.ipc.HBaseClient.getConnection(HBaseClient.java:1002) at org.apache.hadoop.hbase.ipc.HBaseClient.call(HBaseClient.java:854) at org.apache.hadoop.hbase.ipc.WritableRpcEngine$Invoker.invoke(WritableRpcEngine.java:148) at $Proxy9.openRegion(Unknown Source) at org.apache.hadoop.hbase.master.ServerManager.sendRegionOpen(ServerManager.java:407) at org.apache.hadoop.hbase.master.AssignmentManager.assign(AssignmentManager.java:1408) at org.apache.hadoop.hbase.master.AssignmentManager.assign(AssignmentManager.java:1153) at org.apache.hadoop.hbase.master.AssignmentManager.assign(AssignmentManager.java:1128) at org.apache.hadoop.hbase.master.AssignmentManager.assign(AssignmentManager.java:1123) at org.apache.hadoop.hbase.master.AssignmentManager.assignRoot(AssignmentManager.java:1788) at org.apache.hadoop.hbase.master.handler.ServerShutdownHandler.verifyAndAssignRoot(ServerShutdownHandler.java:100) at org.apache.hadoop.hbase.master.handler.ServerShutdownHandler.verifyAndAssignRootWithRetries(ServerShutdownHandler.java:118) at org.apache.hadoop.hbase.master.handler.ServerShutdownHandler.process(ServerShutdownHandler.java:181) at org.apache.hadoop.hbase.executor.EventHandler.run(EventHandler.java:167) at java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) at java.lang.Thread.run(Thread.java:662) 2011-09-13 13:25:52,743 WARN org.apache.hadoop.hbase.master.AssignmentManager: Unable to find a viable location to assign region -ROOT-,,0.70236052 Possible fixes: 1. Have serverManager handle server online event similar to how RegionServerTracker.java calls servermanager.expireServer in the case server goes down. 2. Make timeoutMonitor handle the situation better. This is a special situation in the cluster. 30 minutes timeout can be skipped. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-2834) Deferred deletes
[ https://issues.apache.org/jira/browse/HBASE-2834?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179849#comment-13179849 ] stack commented on HBASE-2834: -- We can close this issue then? (After adding a bit of 'how to do deferred deletes' to the manual?) Deferred deletes Key: HBASE-2834 URL: https://issues.apache.org/jira/browse/HBASE-2834 Project: HBase Issue Type: New Feature Reporter: Andrew Purtell Tangentally mentioned in a blog post, James Hamilton talks about deferred deletes: {quote} If you have an application error, administrative error, or database implementation bug that losses data, then it is simply gone unless you have an offline copy. This, by the way, is why I'm a big fan of deferred delete. This is a technique where deleted items are marked as deleted but not garbage collected until some days or preferably weeks later. Deferred delete is not full protection but it has saved my butt more than once and I'm a believer. See On Designing and Deploying Internet-Scale Services (http://mvdirona.com/jrh/talksAndPapers/JamesRH_Lisa.pdf) for more detail. {quote} (See http://perspectives.mvdirona.com/2010/04/07/StonebrakerOnCAPTheoremAndDatabases.aspx) Because deletes -- at least, after the initial write has been flushed from memstore -- are tombstones, deferred delete in HBase could be supported if somehow tombstones could be invalidated, an undelete operation in effect. This could be accomplished by adding support for tombstones for deletes. Would complicate major compaction but otherwise not touch much. A typical use case might be resurrect any data deleted from _ts1_ to _ts2_ , a period of 4 hours when an application error was operative. In this case a new write would be issued to the table that is a tombstone covering any deletes over that period of time. Users would defer major compactions until safe checkpoint periods. Such guarantees could optionally be extended to the memstoe by using tombstones there as well. But it would probably be sufficient to provide guidance that forcing a flush is necessary to insure edits are persisted in a way that allows for undeletion. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5125) Upgrade hadoop to 1.0.0
Upgrade hadoop to 1.0.0 --- Key: HBASE-5125 URL: https://issues.apache.org/jira/browse/HBASE-5125 Project: HBase Issue Type: Task Reporter: stack Fix For: 0.92.0 Attachments: 5125.txt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5125) Upgrade hadoop to 1.0.0
[ https://issues.apache.org/jira/browse/HBASE-5125?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-5125: - Attachment: 5125.txt Upgrade hadoop to 1.0.0 --- Key: HBASE-5125 URL: https://issues.apache.org/jira/browse/HBASE-5125 Project: HBase Issue Type: Task Reporter: stack Fix For: 0.92.0 Attachments: 5125.txt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5125) Upgrade hadoop to 1.0.0
[ https://issues.apache.org/jira/browse/HBASE-5125?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-5125: - Status: Patch Available (was: Open) Upgrade hadoop to 1.0.0 --- Key: HBASE-5125 URL: https://issues.apache.org/jira/browse/HBASE-5125 Project: HBase Issue Type: Task Reporter: stack Fix For: 0.92.0 Attachments: 5125.txt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HBASE-5093) wiki update for HBase/Scala
[ https://issues.apache.org/jira/browse/HBASE-5093?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Joe Stein resolved HBASE-5093. -- Resolution: Fixed good to go thanks, wiki updated wiki update for HBase/Scala --- Key: HBASE-5093 URL: https://issues.apache.org/jira/browse/HBASE-5093 Project: HBase Issue Type: Improvement Reporter: Joe Stein I tried to edit the wiki but it says immutable page would be helpful/nice for folks to know how to get sbt working with Scala the following is what I did to get it working, not sure why could not edit the wiki figure i open a JIRA so someone with access could update this {code} resolvers += Apache HBase at https://repository.apache.org/content/repositories/releases; resolvers += Thrift at http://people.apache.org/~rawson/repo/; libraryDependencies ++= Seq( org.apache.hadoop % hadoop-core % 0.20.2, org.apache.hbase % hbase % 0.90.4 ) {code} or let me access it and I can do it, np -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179891#comment-13179891 ] Prakash Khemani commented on HBASE-5081: The retry logic is in HMaster.splitLogAfterStartup(). I will remove the OrphanLogException handling from MasterFileSystem. On 1/4/12 12:03 PM, Zhihong Yu (Commented) (JIRA) j...@apache.org Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5041) Major compaction on non existing table does not throw error
[ https://issues.apache.org/jira/browse/HBASE-5041?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179890#comment-13179890 ] Zhihong Yu commented on HBASE-5041: --- TestReplication failed in test suite run. When I ran it separately, I got: {code} Failed tests: queueFailover(org.apache.hadoop.hbase.replication.TestReplication) {code} Looking at the patch for 0.90, it doesn't seem to be related to replication. @J-D: Your review/comment on the patch would be helpful. Thanks Major compaction on non existing table does not throw error Key: HBASE-5041 URL: https://issues.apache.org/jira/browse/HBASE-5041 Project: HBase Issue Type: Bug Components: regionserver, shell Affects Versions: 0.90.3 Reporter: Shrijeet Paliwal Assignee: Shrijeet Paliwal Fix For: 0.92.0, 0.94.0, 0.90.6 Attachments: 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch Following will not complain even if fubar does not exist {code} echo major_compact 'fubar' | $HBASE_HOME/bin/hbase shell {code} The downside for this defect is that major compaction may be skipped due to a typo by Ops. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179894#comment-13179894 ] Prakash Khemani commented on HBASE-5081: Will look into the test failure. I am not sure I know where to find the test run's output logs. On 1/4/12 12:35 PM, Zhihong Yu (Commented) (JIRA) j...@apache.org Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5041) Major compaction on non existing table does not throw error
[ https://issues.apache.org/jira/browse/HBASE-5041?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179907#comment-13179907 ] Shrijeet Paliwal commented on HBASE-5041: - {code} mvn clean compile test -Dtest=TestReplication {code} Above passes without error for branch 0.90 in my dev machine. -Shrijeet Major compaction on non existing table does not throw error Key: HBASE-5041 URL: https://issues.apache.org/jira/browse/HBASE-5041 Project: HBase Issue Type: Bug Components: regionserver, shell Affects Versions: 0.90.3 Reporter: Shrijeet Paliwal Assignee: Shrijeet Paliwal Fix For: 0.92.0, 0.94.0, 0.90.6 Attachments: 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0002-HBASE-5041-Throw-error-if-table-does-not-exist.patch, 0003-HBASE-5041-Throw-error-if-table-does-not-exist.0.90.patch Following will not complain even if fubar does not exist {code} echo major_compact 'fubar' | $HBASE_HOME/bin/hbase shell {code} The downside for this defect is that major compaction may be skipped due to a typo by Ops. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Prakash Khemani updated HBASE-5081: --- Attachment: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch set the interrupt flag on InterruptedException. Remove OrphanLogException handling for distributed log splitting. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179912#comment-13179912 ] Zhihong Yu commented on HBASE-5081: --- bq. If the caller thread is interrupted while waiting for status to change to DELETED ... My comment @ 04/Jan/12 19:46 was about oldtask.wait() being awakened due to spurious wakeup from JVM. You can use the following to mimic the environment of hadoopx build machines when you run unit tests: {code} -argLine-enableassertions -Xmx1900m -Djava.security.egd=file:/dev/./urandom/argLine +argLine-d32 -enableassertions -Xmx1900m -Djava.security.egd=file:/dev/./urandom/argLine {code} Looking at https://builds.apache.org/job/PreCommit-HBASE-Build/665/testReport/org.apache.hadoop.hbase.master/TestSplitLogManager/, it seems all the tests passed. Maybe N Keywal would have some idea about this. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179926#comment-13179926 ] Prakash Khemani commented on HBASE-5081: If there is a spurious wakeup before the status has changed to DELETE then the code will return error (oldtask) to the caller. Regarding the hung TestSplitLogManager test in https://builds.apache.org/job/PreCommit-HBASE-Build/665/console - I couldn't find what failed or what hung. https://builds.apache.org/job/PreCommit-HBASE-Build/665//testReport/org.apa che.hadoop.hbase.master/TestSplitLogManager/ shows that everything passed. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5125) Upgrade hadoop to 1.0.0
[ https://issues.apache.org/jira/browse/HBASE-5125?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179931#comment-13179931 ] Hadoop QA commented on HBASE-5125: -- -1 overall. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12509452/5125.txt against trunk revision . +1 @author. The patch does not contain any @author tags. -1 tests included. The patch doesn't appear to include any new or modified tests. Please justify why no new tests are needed for this patch. Also please list what manual steps were performed to verify this patch. -1 javadoc. The javadoc tool appears to have generated -151 warning messages. +1 javac. The applied patch does not increase the total number of javac compiler warnings. -1 findbugs. The patch appears to introduce 77 new Findbugs (version 1.3.9) warnings. +1 release audit. The applied patch does not increase the total number of release audit warnings. -1 core tests. The patch failed these unit tests: org.apache.hadoop.hbase.mapreduce.TestImportTsv org.apache.hadoop.hbase.mapred.TestTableMapReduce org.apache.hadoop.hbase.mapreduce.TestHFileOutputFormat Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/667//testReport/ Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/667//artifact/trunk/patchprocess/newPatchFindbugsWarnings.html Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/667//console This message is automatically generated. Upgrade hadoop to 1.0.0 --- Key: HBASE-5125 URL: https://issues.apache.org/jira/browse/HBASE-5125 Project: HBase Issue Type: Task Reporter: stack Fix For: 0.92.0 Attachments: 5125.txt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179933#comment-13179933 ] Lars Hofhansl commented on HBASE-2947: -- Good point... I just checked, and Row extends WritableComparable, which in turn extends Writable... So should be good. Maybe than Append can be simplified to only implement Row and not also Writable. MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179935#comment-13179935 ] Hadoop QA commented on HBASE-5081: -- -1 overall. Here are the results of testing the latest attachment http://issues.apache.org/jira/secure/attachment/12509461/0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch against trunk revision . +1 @author. The patch does not contain any @author tags. +1 tests included. The patch appears to include 8 new or modified tests. -1 javadoc. The javadoc tool appears to have generated -151 warning messages. +1 javac. The applied patch does not increase the total number of javac compiler warnings. -1 findbugs. The patch appears to introduce 77 new Findbugs (version 1.3.9) warnings. +1 release audit. The applied patch does not increase the total number of release audit warnings. +1 core tests. The patch passed unit tests in . Test results: https://builds.apache.org/job/PreCommit-HBASE-Build/668//testReport/ Findbugs warnings: https://builds.apache.org/job/PreCommit-HBASE-Build/668//artifact/trunk/patchprocess/newPatchFindbugsWarnings.html Console output: https://builds.apache.org/job/PreCommit-HBASE-Build/668//console This message is automatically generated. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179938#comment-13179938 ] stack commented on HBASE-5120: -- When would it ever make sense that TimeoutMonitor ask a regionserver RE-close a region? It implies that somehow the regionserver 'forgot' to do the close or that it dropped the message. If either happened, we have bigger problems and having the TM do a new close will only make it worse (If a RS 'forgot' or 'failed' a close, its probably on its way out and its ephemeral node is about to vanish and master shutdown handler will do the close fixups; if it does not work this way, we should make it so?). TM might make some sense for region openings but even here, I'd think the above applies; i.e. if a RS can't open a region, it should kill itself. If the region is bad, its unlikely any RS will open it unless the error transient. For this latter case the TM running every 30 minutes, while conservative, is probably fine for a default. I think we should leave the TM at its current 30 minute timeout and undo this issue as critical against 0.92. There are outstanding race conditions in hbase around shutdown handler and master operations that deserve more attention that this IMO. Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Fix For: 0.92.0, 0.94.0 Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT
[jira] [Commented] (HBASE-5120) Timeout monitor races with table disable handler
[ https://issues.apache.org/jira/browse/HBASE-5120?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179940#comment-13179940 ] stack commented on HBASE-5120: -- On the patch, Ram, should we even try unassign if server in this.regions is null; i.e. we should fail way earlier? IMO, this AM needs a rewrite if only to make its various transitions testable. There are also too many methods doing similar things with lots of overlap; its really hard to figure whats going on. Timeout monitor races with table disable handler Key: HBASE-5120 URL: https://issues.apache.org/jira/browse/HBASE-5120 Project: HBase Issue Type: Bug Affects Versions: 0.92.0 Reporter: Zhihong Yu Priority: Blocker Fix For: 0.92.0, 0.94.0 Attachments: HBASE-5120.patch Here is what J-D described here: https://issues.apache.org/jira/browse/HBASE-5119?focusedCommentId=13179176page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-13179176 I think I will retract from my statement that it used to be extremely racy and caused more troubles than it fixed, on my first test I got a stuck region in transition instead of being able to recover. The timeout was set to 2 minutes to be sure I hit it. First the region gets closed {quote} 2012-01-04 00:16:25,811 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Sent CLOSE to sv4r5s38,62023,1325635980913 for region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. {quote} 2 minutes later it times out: {quote} 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Regions in transition timed out: test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636185810, server=null 2012-01-04 00:18:30,026 INFO org.apache.hadoop.hbase.master.AssignmentManager: Region has been PENDING_CLOSE for too long, running forced unassign again on region=test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,027 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Starting unassignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. (offlining) {quote} 100ms later the master finally gets the event: {quote} 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Handling transition=RS_ZK_REGION_CLOSED, server=sv4r5s38,62023,1325635980913, region=1a4b111bcc228043e89f59c4c3f6a791, which is more than 15 seconds late 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.handler.ClosedRegionHandler: Handling CLOSED event for 1a4b111bcc228043e89f59c4c3f6a791 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: Table being disabled so deleting ZK node and removing from regions in transition, skipping assignment of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. 2012-01-04 00:18:30,129 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Deleting existing unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 that is in expected state RS_ZK_REGION_CLOSED 2012-01-04 00:18:30,166 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Successfully deleted unassigned node for region 1a4b111bcc228043e89f59c4c3f6a791 in expected state RS_ZK_REGION_CLOSED {quote} At this point everything is fine, the region was processed as closed. But wait, remember that line where it said it was going to force an unassign? {quote} 2012-01-04 00:18:30,322 DEBUG org.apache.hadoop.hbase.zookeeper.ZKAssign: master:62003-0x134589d3db03587 Creating unassigned node for 1a4b111bcc228043e89f59c4c3f6a791 in a CLOSING state 2012-01-04 00:18:30,328 INFO org.apache.hadoop.hbase.master.AssignmentManager: Server null returned java.lang.NullPointerException: Passed server is null for 1a4b111bcc228043e89f59c4c3f6a791 {quote} Now the master is confused, it recreated the RIT znode but the region doesn't even exist anymore. It even tries to shut it down but is blocked by NPEs. Now this is what's going on. The late ZK notification that the znode was deleted (but it got recreated after): {quote} 2012-01-04 00:19:33,285 DEBUG org.apache.hadoop.hbase.master.AssignmentManager: The znode of region test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. has been deleted. {quote} Then it prints this, and much later tries to unassign it again: {quote} 2012-01-04 00:19:46,607 DEBUG org.apache.hadoop.hbase.master.handler.DeleteTableHandler: Waiting on region to clear regions in transition; test1,089cd0c9,1325635015491.1a4b111bcc228043e89f59c4c3f6a791. state=PENDING_CLOSE, ts=1325636310328, server=null ... 2012-01-04
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179945#comment-13179945 ] Mikhail Bautin commented on HBASE-4218: --- I think that with an 8K line patch we probably should not try to put more complexity into the first version of delta encoding. We can always make things more complicated later. I like the two-parameter setup: DATA_BLOCK_ENCODING sets the encoding type (on-disk and in-cache by default) and ENCODE_ON_DISK (true by default) allows to use in-cache-only encoding (when explicitly setting ENCODE_ON_DISK=false) and get the benefit of encoding in cache even before we are 100% sure that our encoding algorithms and encoded scanners are stable. If everyone agrees with that, I will finish the patch by (1) adding a unit test for switching data block encoding column family settings; (2) including encoding type in the cache key; and (3) simplifying the HFileDataBlockEncoder interface, since we assume that the in-memory format (used by scanners) is always the same as the in-cache format and don't need methods such as afterReadFromDiskAndPuttingInCache anymore. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179944#comment-13179944 ] stack commented on HBASE-5081: -- I'm running tests locally. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5125) Upgrade hadoop to 1.0.0
[ https://issues.apache.org/jira/browse/HBASE-5125?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] stack updated HBASE-5125: - Resolution: Fixed Assignee: stack Status: Resolved (was: Patch Available) Committed 0.92 and trunk Upgrade hadoop to 1.0.0 --- Key: HBASE-5125 URL: https://issues.apache.org/jira/browse/HBASE-5125 Project: HBase Issue Type: Task Reporter: stack Assignee: stack Fix For: 0.92.0 Attachments: 5125.txt -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179955#comment-13179955 ] Matt Corgan commented on HBASE-4218: {quote}I think that with an 8K line patch we probably should not try to put more complexity into the first version of delta encoding.{quote}Yes, totally agreeing here. It is a work in progress, and so these settings in this patch don't have to make perfect sense. I like the latest DATA_BLOCK_ENCODING=NONE(?) and ENCODE_ON_DISK=true defaults. All other comments look sensible. Have you covered the case where you have encoded blocks in the block cache and are compacting to an unencoded hfile? You will want to make sure that you are using (not ignoring) the cached blocks. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179960#comment-13179960 ] Mikhail Bautin commented on HBASE-4218: --- Re-reading my previous post, I want to make an addition: we still use cached encoded blocks when compacting a fully-encoded column family. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179959#comment-13179959 ] Mikhail Bautin commented on HBASE-4218: --- Actually, I think it is OK to ignore cached encoded blocks on compaction. We can get encoded blocks in cache and have a compaction write an unencoded file in two cases: * Encoding is turned on in cache only. In that case we don't want to use encoded blocks during compaction at all, because the in-cache-only mode implies that we don't trust our encoding algorithms 100% and want to guard against possible persistent data corruption. * Encoding was turned on (either in cache only or everywhere) and it was turned off entirely. Since this is not a very frequent case, I think we could probably optimize this after the patch is stabilized. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Lars Hofhansl updated HBASE-2947: - Attachment: 2947-final.txt Patch that I committed (based on Stack's comment Append does not actually need to implement Writable, since Row already extends Writable). MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-final.txt, 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-2947) MultiIncrement/MultiAppend (MultiGet functionality for increments and appends)
[ https://issues.apache.org/jira/browse/HBASE-2947?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Lars Hofhansl updated HBASE-2947: - Resolution: Fixed Hadoop Flags: Reviewed Status: Resolved (was: Patch Available) Thanks for the review Stack! MultiIncrement/MultiAppend (MultiGet functionality for increments and appends) -- Key: HBASE-2947 URL: https://issues.apache.org/jira/browse/HBASE-2947 Project: HBase Issue Type: New Feature Components: client, regionserver Reporter: Jonathan Gray Assignee: Lars Hofhansl Priority: Minor Fix For: 0.94.0 Attachments: 2947-final.txt, 2947-v2.txt, HBASE-2947-v1.patch HBASE-1845 introduced MultiGet and other cross-row/cross-region batch operations. We should add a way to do that with increments. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-4218) Data Block Encoding of KeyValues (aka delta encoding / prefix compression)
[ https://issues.apache.org/jira/browse/HBASE-4218?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179973#comment-13179973 ] Matt Corgan commented on HBASE-4218: {quote}I think it is OK to ignore cached encoded blocks on compaction{quote}The circumstance i was worried about is if you are doing many small flushes and minor compactions. The blocks to be compacted could mostly be in cache, and you would be ignoring them all. I guess it doesn't matter if it's just for testing, but might give a false impression of performance. Data Block Encoding of KeyValues (aka delta encoding / prefix compression) --- Key: HBASE-4218 URL: https://issues.apache.org/jira/browse/HBASE-4218 Project: HBase Issue Type: Improvement Components: io Affects Versions: 0.94.0 Reporter: Jacek Migdal Assignee: Mikhail Bautin Labels: compression Fix For: 0.94.0 Attachments: 0001-Delta-encoding-fixed-encoded-scanners.patch, 0001-Delta-encoding.patch, 4218-v16.txt, 4218.txt, D447.1.patch, D447.10.patch, D447.11.patch, D447.12.patch, D447.13.patch, D447.14.patch, D447.15.patch, D447.16.patch, D447.17.patch, D447.2.patch, D447.3.patch, D447.4.patch, D447.5.patch, D447.6.patch, D447.7.patch, D447.8.patch, D447.9.patch, Data-block-encoding-2011-12-23.patch, Delta-encoding.patch-2011-12-22_11_52_07.patch, Delta_encoding_with_memstore_TS.patch, open-source.diff A compression for keys. Keys are sorted in HFile and they are usually very similar. Because of that, it is possible to design better compression than general purpose algorithms, It is an additional step designed to be used in memory. It aims to save memory in cache as well as speeding seeks within HFileBlocks. It should improve performance a lot, if key lengths are larger than value lengths. For example, it makes a lot of sense to use it when value is a counter. Initial tests on real data (key length = ~ 90 bytes , value length = 8 bytes) shows that I could achieve decent level of compression: key compression ratio: 92% total compression ratio: 85% LZO on the same data: 85% LZO after delta encoding: 91% While having much better performance (20-80% faster decompression ratio than LZO). Moreover, it should allow far more efficient seeking which should improve performance a bit. It seems that a simple compression algorithms are good enough. Most of the savings are due to prefix compression, int128 encoding, timestamp diffs and bitfields to avoid duplication. That way, comparisons of compressed data can be much faster than a byte comparator (thanks to prefix compression and bitfields). In order to implement it in HBase two important changes in design will be needed: -solidify interface to HFileBlock / HFileReader Scanner to provide seeking and iterating; access to uncompressed buffer in HFileBlock will have bad performance -extend comparators to support comparison assuming that N first bytes are equal (or some fields are equal) Link to a discussion about something similar: http://search-hadoop.com/m/5aqGXJEnaD1/hbase+windowssubj=Re+prefix+compression -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter1); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Kannan Muthukkaruppan updated HBASE-5126: - Attachment: testAndTwoPrefixFilters.rb full test attached. AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter1); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (HBASE-5031) [89-fb] Remove hard-coded non-existent host name from TestScanner
[ https://issues.apache.org/jira/browse/HBASE-5031?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Nicolas Spiegelberg resolved HBASE-5031. Resolution: Fixed Release Note: Fixed in 0.89-fb. Not necessary in trunk [89-fb] Remove hard-coded non-existent host name from TestScanner -- Key: HBASE-5031 URL: https://issues.apache.org/jira/browse/HBASE-5031 Project: HBase Issue Type: Bug Reporter: Mikhail Bautin Priority: Minor Attachments: D867.1.patch TestScanner is failing on 0.89-fb because it has a hard-coded fake host name that it is trying to look up. Replacing this with 127.0.0.1:random_port instead. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13179991#comment-13179991 ] Madhuwanti Vaidya commented on HBASE-5126: -- Kannan: You added filter1 twice. AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter1); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5124) Backport LoadTestTool to 0.92
[ https://issues.apache.org/jira/browse/HBASE-5124?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Yu updated HBASE-5124: -- Attachment: hbase-5124.txt The patch, v1 Backport LoadTestTool to 0.92 - Key: HBASE-5124 URL: https://issues.apache.org/jira/browse/HBASE-5124 Project: HBase Issue Type: Task Reporter: Zhihong Yu Fix For: 0.92.0 Attachments: hbase-5124.txt LoadTestTool is very useful. This JIRA backports LoadTestTool to 0.92 so that users don't have to build TRUNK in order to use it against 0.92 cluster. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (HBASE-5124) Backport LoadTestTool to 0.92
[ https://issues.apache.org/jira/browse/HBASE-5124?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Yu reassigned HBASE-5124: - Assignee: Zhihong Yu Backport LoadTestTool to 0.92 - Key: HBASE-5124 URL: https://issues.apache.org/jira/browse/HBASE-5124 Project: HBase Issue Type: Task Reporter: Zhihong Yu Assignee: Zhihong Yu Fix For: 0.92.0 Attachments: hbase-5124.txt LoadTestTool is very useful. This JIRA backports LoadTestTool to 0.92 so that users don't have to build TRUNK in order to use it against 0.92 cluster. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=1318#comment-1318 ] stack commented on HBASE-5081: -- All tests but TestClassLoading (unrelated) pass for me locally: {code} ... Results : Failed tests: testClassLoadingFromHDFS(org.apache.hadoop.hbase.coprocessor.TestClassLoading): Class TestCP1 was missing on a region Tests in error: testEnableTableRoundRobinAssignment(org.apache.hadoop.hbase.client.TestAdmin): org.apache.hadoop.hbase.TableNotEnabledException: testEnableTableAssignment Tests run: 792, Failures: 1, Errors: 1, Skipped: 10 [INFO] [ERROR] BUILD FAILURE [INFO] [INFO] There are test failures. {code} Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13180002#comment-13180002 ] stack commented on HBASE-5081: -- You down w/ my committing Ted? Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13180008#comment-13180008 ] Amitanand Aiyer commented on HBASE-5126: Fixed the bug in the test. But the issue still persists. There are rows that get printed. AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter1); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13180018#comment-13180018 ] stack commented on HBASE-5081: -- Put up your version of the patch Ted and lets commit it. I'll roll a new RC and lets test that rather than individual patch. Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Kannan Muthukkaruppan updated HBASE-5126: - Attachment: testAndTwoPrefixFilters.rb AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb, testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter2); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Kannan Muthukkaruppan updated HBASE-5126: - Description: [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter2); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} was: [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter1); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb, testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter2); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (HBASE-5127) [ref manual] Better block cache documentation
[ref manual] Better block cache documentation - Key: HBASE-5127 URL: https://issues.apache.org/jira/browse/HBASE-5127 Project: HBase Issue Type: Task Reporter: Jean-Daniel Cryans Assignee: Jean-Daniel Cryans I've been playing a lot with 0.92's block caching and I wrote down some documentation that I think will be useful to others. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5127) [ref manual] Better block cache documentation
[ https://issues.apache.org/jira/browse/HBASE-5127?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jean-Daniel Cryans updated HBASE-5127: -- Attachment: HBASE-5127.patch Here's the patch for book.xml, anyone can review both the content and my english? [ref manual] Better block cache documentation - Key: HBASE-5127 URL: https://issues.apache.org/jira/browse/HBASE-5127 Project: HBase Issue Type: Task Reporter: Jean-Daniel Cryans Assignee: Jean-Daniel Cryans Attachments: HBASE-5127.patch I've been playing a lot with 0.92's block caching and I wrote down some documentation that I think will be useful to others. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (HBASE-5081) Distributed log splitting deleteNode races against splitLog retry
[ https://issues.apache.org/jira/browse/HBASE-5081?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zhihong Yu updated HBASE-5081: -- Attachment: 5081-deleteNode-with-while-loop.txt Patch based on Prakash's latest patch. Changed 'if (oldtask.status == FAILURE)' to a while loop. Also restored @param for the first SplitLogManager ctor Distributed log splitting deleteNode races against splitLog retry -- Key: HBASE-5081 URL: https://issues.apache.org/jira/browse/HBASE-5081 Project: HBase Issue Type: Bug Components: wal Affects Versions: 0.92.0, 0.94.0 Reporter: Jimmy Xiang Assignee: Prakash Khemani Fix For: 0.92.0 Attachments: 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 0001-HBASE-5081-jira-Distributed-log-splitting-deleteNode.patch, 5081-deleteNode-with-while-loop.txt, distributed-log-splitting-screenshot.png, hbase-5081-patch-v6.txt, hbase-5081-patch-v7.txt, hbase-5081_patch_for_92_v4.txt, hbase-5081_patch_v5.txt, patch_for_92.txt, patch_for_92_v2.txt, patch_for_92_v3.txt Recently, during 0.92 rc testing, we found distributed log splitting hangs there forever. Please see attached screen shot. I looked into it and here is what happened I think: 1. One rs died, the servershutdownhandler found it out and started the distributed log splitting; 2. All three tasks failed, so the three tasks were deleted, asynchronously; 3. Servershutdownhandler retried the log splitting; 4. During the retrial, it created these three tasks again, and put them in a hashmap (tasks); 5. The asynchronously deletion in step 2 finally happened for one task, in the callback, it removed one task in the hashmap; 6. One of the newly submitted tasks' zookeeper watcher found out that task is unassigned, and it is not in the hashmap, so it created a new orphan task. 7. All three tasks failed, but that task created in step 6 is an orphan so the batch.err counter was one short, so the log splitting hangs there and keeps waiting for the last task to finish which is never going to happen. So I think the problem is step 2. The fix is to make deletion sync, instead of async, so that the retry will have a clean start. Async deleteNode will mess up with split log retrial. In extreme situation, if async deleteNode doesn't happen soon enough, some node created during the retrial could be deleted. deleteNode should be sync. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HBASE-5126) AND (USING FilterList) of two ColumnPrefixFilters broken
[ https://issues.apache.org/jira/browse/HBASE-5126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13180054#comment-13180054 ] Amitanand Aiyer commented on HBASE-5126: Seems like this is fixed in trunk ... although they also have new functions transform that doesn't exist in the 89 branch. AND (USING FilterList) of two ColumnPrefixFilters broken Key: HBASE-5126 URL: https://issues.apache.org/jira/browse/HBASE-5126 Project: HBase Issue Type: Bug Reporter: Kannan Muthukkaruppan Attachments: testAndTwoPrefixFilters.rb, testAndTwoPrefixFilters.rb [Notice this in 89 branch. Possibly an issue in trunk also.] A test which does a columnPrefixFilter(tag0) AND columnPrefixFilter(tag1) should return 0 kvs; instead it returns kvs with prefix tag0. {code} table = HTable.new(conf, tableName) put = Put.new(Bytes.toBytes(row)) put.add(cf1name, Bytes.toBytes(tag0), Bytes.toBytes(value0)) put.add(cf1name, Bytes.toBytes(tag1), Bytes.toBytes(value1)) put.add(cf1name, Bytes.toBytes(tag2), Bytes.toBytes(value2)) table.put(put) # Test for AND Two Column Prefix Filters filter1 = ColumnPrefixFilter.new(Bytes.toBytes(tag0)); filter2 = ColumnPrefixFilter.new(Bytes.toBytes(tag2)); filters = FilterList.new(FilterList::Operator::MUST_PASS_ALL); filters.addFilter(filter1); filters.addFilter(filter2); get = Get.new(Bytes.toBytes(row)) get.setFilter(filters) get.setMaxVersions(); keyValues = table.get(get).raw() keyValues.each do |keyValue| puts Key=#{Bytes.toStringBinary(keyValue.getQualifier())}; Value=#{Bytes.toStringBinary(keyValue.getValue())}; Timestamp=#{keyValue.getTimestamp()} end {code} outputs: {code} Key=tag0; Value=value0; Timestamp=1325719223523 {code} -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (HBASE-3949) Add Master link to RegionServer pages
[ https://issues.apache.org/jira/browse/HBASE-3949?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Todd Lipcon reassigned HBASE-3949: -- Assignee: Gregory Chanan Add Master link to RegionServer pages --- Key: HBASE-3949 URL: https://issues.apache.org/jira/browse/HBASE-3949 Project: HBase Issue Type: Improvement Components: regionserver Affects Versions: 0.90.3, 0.92.0 Reporter: Lars George Assignee: Gregory Chanan Priority: Minor Labels: noob Use the ZK info where the master is to add a UI link on the top of each RegionServer page. Currently you cannot navigate directly to the Master UI once you are on a RS page. Not sure if the info port is exposed OTTOMH, but we could either use the RS local config setting for that or add it to ZK to enable lookup. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa For more information on JIRA, see: http://www.atlassian.com/software/jira