[jira] [Commented] (HDFS-8792) Improve BlockManager#postponedMisreplicatedBlocks and BlockManager#excessReplicateMap
[ https://issues.apache.org/jira/browse/HDFS-8792?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647537#comment-14647537 ] Hadoop QA commented on HDFS-8792: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem || Runtime || Comment || | {color:red}-1{color} | pre-patch | 15m 17s | Findbugs (version ) appears to be broken on trunk. | | {color:green}+1{color} | @author | 0m 0s | The patch does not contain any @author tags. | | {color:green}+1{color} | tests included | 0m 0s | The patch appears to include 1 new or modified test files. | | {color:green}+1{color} | javac | 7m 40s | There were no new javac warning messages. | | {color:green}+1{color} | javadoc | 9m 42s | There were no new javadoc warning messages. | | {color:green}+1{color} | release audit | 0m 22s | The applied patch does not increase the total number of release audit warnings. | | {color:green}+1{color} | checkstyle | 0m 36s | There were no new checkstyle issues. | | {color:red}-1{color} | whitespace | 0m 0s | The patch has 3 line(s) that end in whitespace. Use git apply --whitespace=fix. | | {color:green}+1{color} | install | 1m 29s | mvn install still works. | | {color:green}+1{color} | eclipse:eclipse | 0m 32s | The patch built with eclipse:eclipse. | | {color:green}+1{color} | findbugs | 2m 28s | The patch does not introduce any new Findbugs (version 3.0.0) warnings. | | {color:green}+1{color} | native | 3m 3s | Pre-build of native portion | | {color:red}-1{color} | hdfs tests | 165m 14s | Tests failed in hadoop-hdfs. | | | | 206m 25s | | \\ \\ || Reason || Tests || | Failed unit tests | hadoop.hdfs.server.namenode.snapshot.TestRenameWithSnapshots | \\ \\ || Subsystem || Report/Notes || | Patch URL | http://issues.apache.org/jira/secure/attachment/12747948/HDFS-8792.002.patch | | Optional Tests | javadoc javac unit findbugs checkstyle | | git revision | trunk / ddc867ce | | whitespace | https://builds.apache.org/job/PreCommit-HDFS-Build/11874/artifact/patchprocess/whitespace.txt | | hadoop-hdfs test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11874/artifact/patchprocess/testrun_hadoop-hdfs.txt | | Test Results | https://builds.apache.org/job/PreCommit-HDFS-Build/11874/testReport/ | | Java | 1.7.0_55 | | uname | Linux asf904.gq1.ygridcore.net 3.13.0-36-lowlatency #63-Ubuntu SMP PREEMPT Wed Sep 3 21:56:12 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux | | Console output | https://builds.apache.org/job/PreCommit-HDFS-Build/11874/console | This message was automatically generated. Improve BlockManager#postponedMisreplicatedBlocks and BlockManager#excessReplicateMap - Key: HDFS-8792 URL: https://issues.apache.org/jira/browse/HDFS-8792 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Yi Liu Assignee: Yi Liu Attachments: HDFS-8792.001.patch, HDFS-8792.002.patch {{LightWeightHashSet}} requires fewer memory than java hashset. Furthermore, for {{excessReplicateMap}}, we can use {{HashMap}} instead of {{TreeMap}} instead, since no need to sort. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8202) Improve end to end stirpping file test to add erasure recovering test
[ https://issues.apache.org/jira/browse/HDFS-8202?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xinwei Qin updated HDFS-8202: -- Attachment: HDFS-8202-HDFS-7285.006.patch Thanks [~zhz] for the comments. Upload 006 patch for review. bq.TestWriteStripedFileWithFailure actually fails, could you debug the issue? This issue is addressed by HDFS-8704, now, all the test can pass with the latest 003 patch in HDFS-8704. Improve end to end stirpping file test to add erasure recovering test - Key: HDFS-8202 URL: https://issues.apache.org/jira/browse/HDFS-8202 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Kai Zheng Assignee: Xinwei Qin Attachments: HDFS-8202-HDFS-7285.003.patch, HDFS-8202-HDFS-7285.004.patch, HDFS-8202-HDFS-7285.005.patch, HDFS-8202-HDFS-7285.006.patch, HDFS-8202.001.patch, HDFS-8202.002.patch This to follow on HDFS-8201 to add erasure recovering test in the end to end stripping file test: * After writing certain blocks to the test file, delete some block file; * Read the file content and compare, see if any recovering issue, or verify the erasure recovering works or not. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647579#comment-14647579 ] Hudson commented on HDFS-8816: -- FAILURE: Integrated in Hadoop-Hdfs-trunk-Java8 #261 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk-Java8/261/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-6682) Add a metric to expose the timestamp of the oldest under-replicated block
[ https://issues.apache.org/jira/browse/HDFS-6682?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647666#comment-14647666 ] Yi Liu commented on HDFS-6682: -- {quote} There's a big advantage to having this metric: it's extremely useful to know how backlogged the replication queue is as a determinant of namenode health on extremely large clusters. {quote} We have many ways to know about namenode health or in heavy load. It's not worth. Thanks Akira for reverting. Add a metric to expose the timestamp of the oldest under-replicated block - Key: HDFS-6682 URL: https://issues.apache.org/jira/browse/HDFS-6682 Project: Hadoop HDFS Issue Type: Improvement Reporter: Akira AJISAKA Assignee: Akira AJISAKA Labels: metrics Attachments: HDFS-6682.002.patch, HDFS-6682.003.patch, HDFS-6682.004.patch, HDFS-6682.005.patch, HDFS-6682.006.patch, HDFS-6682.patch In the following case, the data in the HDFS is lost and a client needs to put the same file again. # A Client puts a file to HDFS # A DataNode crashes before replicating a block of the file to other DataNodes I propose a metric to expose the timestamp of the oldest under-replicated/corrupt block. That way client can know what file to retain for the re-try. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647498#comment-14647498 ] Hudson commented on HDFS-8816: -- FAILURE: Integrated in Hadoop-Yarn-trunk-Java8 #272 (See [https://builds.apache.org/job/Hadoop-Yarn-trunk-Java8/272/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647784#comment-14647784 ] Hudson commented on HDFS-8816: -- SUCCESS: Integrated in Hadoop-Mapreduce-trunk-Java8 #269 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk-Java8/269/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647825#comment-14647825 ] Hudson commented on HDFS-8816: -- SUCCESS: Integrated in Hadoop-Mapreduce-trunk #2218 (See [https://builds.apache.org/job/Hadoop-Mapreduce-trunk/2218/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-6682) Add a metric to expose the timestamp of the oldest under-replicated block
[ https://issues.apache.org/jira/browse/HDFS-6682?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647679#comment-14647679 ] Allen Wittenauer commented on HDFS-6682: bq. We have many ways to know about namenode health or in heavy load. It's not worth. They don't work for this use case. We see this *every day*. NN is healthy *except* the replication queue is backed up. Add a metric to expose the timestamp of the oldest under-replicated block - Key: HDFS-6682 URL: https://issues.apache.org/jira/browse/HDFS-6682 Project: Hadoop HDFS Issue Type: Improvement Reporter: Akira AJISAKA Assignee: Akira AJISAKA Labels: metrics Attachments: HDFS-6682.002.patch, HDFS-6682.003.patch, HDFS-6682.004.patch, HDFS-6682.005.patch, HDFS-6682.006.patch, HDFS-6682.patch In the following case, the data in the HDFS is lost and a client needs to put the same file again. # A Client puts a file to HDFS # A DataNode crashes before replicating a block of the file to other DataNodes I propose a metric to expose the timestamp of the oldest under-replicated/corrupt block. That way client can know what file to retain for the re-try. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647693#comment-14647693 ] Hudson commented on HDFS-8816: -- SUCCESS: Integrated in Hadoop-Hdfs-trunk #2199 (See [https://builds.apache.org/job/Hadoop-Hdfs-trunk/2199/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (HDFS-8784) BlockInfo#numNodes should be numStorages
[ https://issues.apache.org/jira/browse/HDFS-8784?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Jagadesh Kiran N reassigned HDFS-8784: -- Assignee: Jagadesh Kiran N (was: kanaka kumar avvaru) BlockInfo#numNodes should be numStorages Key: HDFS-8784 URL: https://issues.apache.org/jira/browse/HDFS-8784 Project: Hadoop HDFS Issue Type: Improvement Components: namenode Affects Versions: 2.7.1 Reporter: Zhe Zhang Assignee: Jagadesh Kiran N The method actually returns the number of storages holding a block. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8653) Code cleanup for DatanodeManager, DatanodeDescriptor and DatanodeStorageInfo
[ https://issues.apache.org/jira/browse/HDFS-8653?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648318#comment-14648318 ] Tsz Wo Nicholas Sze commented on HDFS-8653: --- Which JIRA(s ) in HDFS-7285 the patch bases on? Code cleanup for DatanodeManager, DatanodeDescriptor and DatanodeStorageInfo Key: HDFS-8653 URL: https://issues.apache.org/jira/browse/HDFS-8653 Project: Hadoop HDFS Issue Type: Improvement Affects Versions: 2.7.0 Reporter: Zhe Zhang Assignee: Zhe Zhang Fix For: 2.8.0 Attachments: HDFS-8653.00.patch While updating the {{blockmanagement}} module to distribute erasure coding recovery work to Datanode, the HDFS-7285 branch also did some code cleanup that should be merged into trunk independently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8835) Convert BlockInfoUnderConstruction as an interface
[ https://issues.apache.org/jira/browse/HDFS-8835?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648157#comment-14648157 ] Tsz Wo Nicholas Sze commented on HDFS-8835: --- ..., which are not directly related to EC. Pushing those changes to trunk separately will make ease merge reviewing. ... It probably won't help much since people also need to review the patch committed to trunk. Patches got committed to trunk neither means that everyone already has understood the code, nor that the patches make the code simpler and easier.Quite a few people told me that the recent change of HDFS-8487 does make the code harder to understand. It makes the familiar code unfamiliar. Convert BlockInfoUnderConstruction as an interface -- Key: HDFS-8835 URL: https://issues.apache.org/jira/browse/HDFS-8835 Project: Hadoop HDFS Issue Type: Improvement Components: namenode Affects Versions: 2.7.1 Reporter: Zhe Zhang Assignee: Zhe Zhang Per discussion under HDFS-8499, this JIRA aims to convert {{BlockInfoUnderConstruction}} as an interface and {{BlockInfoContiguousUnderConstruction}} as its implementation. The HDFS-7285 branch will add {{BlockInfoStripedUnderConstruction}} as another implementation. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-7192) DN should ignore lazyPersist hint if the writer is not local
[ https://issues.apache.org/jira/browse/HDFS-7192?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648262#comment-14648262 ] Hudson commented on HDFS-7192: -- FAILURE: Integrated in Hadoop-trunk-Commit #8246 (See [https://builds.apache.org/job/Hadoop-trunk-Commit/8246/]) HDFS-7192. DN should ignore lazyPersist hint if the writer is not local. (Contributed by Arpit Agarwal) (arp: rev 88d8736ddeff10a03acaa99a9a0ee99dcfabe590) * hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/datanode/TestDataXceiverLazyPersistHint.java * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataXceiver.java * hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DNConf.java * hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/DFSConfigKeys.java DN should ignore lazyPersist hint if the writer is not local Key: HDFS-7192 URL: https://issues.apache.org/jira/browse/HDFS-7192 Project: Hadoop HDFS Issue Type: Sub-task Components: datanode Reporter: Arpit Agarwal Assignee: Arpit Agarwal Fix For: 2.8.0 Attachments: HDFS-7192.01.patch, HDFS-7192.02.patch, HDFS-7192.03.patch, HDFS-7192.04.patch, HDFS-7192.05.patch The DN should ignore {{allowLazyPersist}} hint to {{DataTransferProtocol#writeBlock}} if the writer is not local. Currently we don't restrict memory writes to local clients. For in-cluster clients this is not an issue as single replica writes default to the local DataNode. But clients outside the cluster can still send this hint. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-7192) DN should ignore lazyPersist hint if the writer is not local
[ https://issues.apache.org/jira/browse/HDFS-7192?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Arpit Agarwal updated HDFS-7192: Resolution: Fixed Hadoop Flags: Reviewed Fix Version/s: 2.8.0 Target Version/s: (was: 2.8.0) Status: Resolved (was: Patch Available) Committed for 2.8.0. DN should ignore lazyPersist hint if the writer is not local Key: HDFS-7192 URL: https://issues.apache.org/jira/browse/HDFS-7192 Project: Hadoop HDFS Issue Type: Sub-task Components: datanode Reporter: Arpit Agarwal Assignee: Arpit Agarwal Fix For: 2.8.0 Attachments: HDFS-7192.01.patch, HDFS-7192.02.patch, HDFS-7192.03.patch, HDFS-7192.04.patch, HDFS-7192.05.patch The DN should ignore {{allowLazyPersist}} hint to {{DataTransferProtocol#writeBlock}} if the writer is not local. Currently we don't restrict memory writes to local clients. For in-cluster clients this is not an issue as single replica writes default to the local DataNode. But clients outside the cluster can still send this hint. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-7192) DN should ignore lazyPersist hint if the writer is not local
[ https://issues.apache.org/jira/browse/HDFS-7192?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Arpit Agarwal updated HDFS-7192: Attachment: HDFS-7192.05.patch Thanks for the review [~xyao]. The Jenkins test failure is unrelated. The v05 patch fixes the checkstyle issue by renaming the parameter. I will commit the v05 patch shortly. Delta wrt v04 patch. {code} - final DataNode datanode, DataChecksum requestedChecksum, + final DataNode dn, DataChecksum requestedChecksum, CachingStrategy cachingStrategy, final boolean allowLazyPersist, final boolean pinning) throws IOException { return new BlockReceiver(block, storageType, in, inAddr, myAddr, stage, newGs, minBytesRcvd, maxBytesRcvd, -clientname, srcDataNode, datanode, requestedChecksum, +clientname, srcDataNode, dn, requestedChecksum, {code} DN should ignore lazyPersist hint if the writer is not local Key: HDFS-7192 URL: https://issues.apache.org/jira/browse/HDFS-7192 Project: Hadoop HDFS Issue Type: Sub-task Components: datanode Reporter: Arpit Agarwal Assignee: Arpit Agarwal Attachments: HDFS-7192.01.patch, HDFS-7192.02.patch, HDFS-7192.03.patch, HDFS-7192.04.patch, HDFS-7192.05.patch The DN should ignore {{allowLazyPersist}} hint to {{DataTransferProtocol#writeBlock}} if the writer is not local. Currently we don't restrict memory writes to local clients. For in-cluster clients this is not an issue as single replica writes default to the local DataNode. But clients outside the cluster can still send this hint. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-742) A down DataNode makes Balancer to hang on repeatingly asking NameNode its partial block list
[ https://issues.apache.org/jira/browse/HDFS-742?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Tsz Wo Nicholas Sze updated HDFS-742: - Status: Patch Available (was: Open) A down DataNode makes Balancer to hang on repeatingly asking NameNode its partial block list Key: HDFS-742 URL: https://issues.apache.org/jira/browse/HDFS-742 Project: Hadoop HDFS Issue Type: Bug Components: balancer mover Reporter: Hairong Kuang Assignee: Mit Desai Priority: Minor Attachments: HDFS-742-trunk.patch, HDFS-742.patch We had a balancer that had not made any progress for a long time. It turned out it was repeatingly asking Namenode for a partial block list of one datanode, which was done while the balancer was running. NameNode should notify Balancer that the datanode is not available and Balancer should stop asking for the datanode's block list. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-742) A down DataNode makes Balancer to hang on repeatingly asking NameNode its partial block list
[ https://issues.apache.org/jira/browse/HDFS-742?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648660#comment-14648660 ] Hadoop QA commented on HDFS-742: \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem || Runtime || Comment || | {color:red}-1{color} | pre-patch | 16m 21s | Findbugs (version ) appears to be broken on trunk. | | {color:green}+1{color} | @author | 0m 0s | The patch does not contain any @author tags. | | {color:red}-1{color} | tests included | 0m 0s | 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. | | {color:green}+1{color} | javac | 7m 58s | There were no new javac warning messages. | | {color:green}+1{color} | javadoc | 9m 55s | There were no new javadoc warning messages. | | {color:green}+1{color} | release audit | 0m 23s | The applied patch does not increase the total number of release audit warnings. | | {color:green}+1{color} | checkstyle | 0m 31s | There were no new checkstyle issues. | | {color:green}+1{color} | whitespace | 0m 0s | The patch has no lines that end in whitespace. | | {color:green}+1{color} | install | 1m 32s | mvn install still works. | | {color:green}+1{color} | eclipse:eclipse | 0m 34s | The patch built with eclipse:eclipse. | | {color:green}+1{color} | findbugs | 2m 34s | The patch does not introduce any new Findbugs (version 3.0.0) warnings. | | {color:green}+1{color} | native | 3m 15s | Pre-build of native portion | | {color:red}-1{color} | hdfs tests | 161m 5s | Tests failed in hadoop-hdfs. | | | | 204m 12s | | \\ \\ || Reason || Tests || | Failed unit tests | hadoop.hdfs.TestDFSUpgradeFromImage | | | hadoop.fs.TestSymlinkHdfsFileSystem | | | hadoop.fs.viewfs.TestViewFileSystemHdfs | \\ \\ || Subsystem || Report/Notes || | Patch URL | http://issues.apache.org/jira/secure/attachment/12747558/HDFS-742-trunk.patch | | Optional Tests | javadoc javac unit findbugs checkstyle | | git revision | trunk / 88d8736 | | hadoop-hdfs test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11875/artifact/patchprocess/testrun_hadoop-hdfs.txt | | Test Results | https://builds.apache.org/job/PreCommit-HDFS-Build/11875/testReport/ | | Java | 1.7.0_55 | | uname | Linux asf903.gq1.ygridcore.net 3.13.0-36-lowlatency #63-Ubuntu SMP PREEMPT Wed Sep 3 21:56:12 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux | | Console output | https://builds.apache.org/job/PreCommit-HDFS-Build/11875/console | This message was automatically generated. A down DataNode makes Balancer to hang on repeatingly asking NameNode its partial block list Key: HDFS-742 URL: https://issues.apache.org/jira/browse/HDFS-742 Project: Hadoop HDFS Issue Type: Bug Components: balancer mover Reporter: Hairong Kuang Assignee: Mit Desai Priority: Minor Attachments: HDFS-742-trunk.patch, HDFS-742.patch We had a balancer that had not made any progress for a long time. It turned out it was repeatingly asking Namenode for a partial block list of one datanode, which was done while the balancer was running. NameNode should notify Balancer that the datanode is not available and Balancer should stop asking for the datanode's block list. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8821) Explain message Operation category X is not supported in state standby
[ https://issues.apache.org/jira/browse/HDFS-8821?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Harsh J updated HDFS-8821: -- Resolution: Fixed Hadoop Flags: Reviewed Fix Version/s: 2.8.0 Status: Resolved (was: Patch Available) Thanks Gautam, test passed locally for me as well. Seems unrelated. I've pushed this into trunk and branch-2; thank you for the continued contributions! Explain message Operation category X is not supported in state standby - Key: HDFS-8821 URL: https://issues.apache.org/jira/browse/HDFS-8821 Project: Hadoop HDFS Issue Type: Improvement Reporter: Gautam Gopalakrishnan Assignee: Gautam Gopalakrishnan Priority: Minor Fix For: 2.8.0 Attachments: HDFS-8821-1.patch, HDFS-8821-2.patch There is one message specifically that causes many users to question the health of their HDFS cluster, namely Operation category READ/WRITE is not supported in state standby. HDFS-3447 is an attempt to lower the logging severity for StandbyException related messages but it is not resolved yet. So this jira is an attempt to explain this particular message so it appears less scary. The text is question 3.17 in the Hadoop Wiki FAQ ref: https://wiki.apache.org/hadoop/FAQ#What_does_the_message_.22Operation_category_READ.2FWRITE_is_not_supported_in_state_standby.22_mean.3F -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-6407) new namenode UI, lost ability to sort columns in datanode tab
[ https://issues.apache.org/jira/browse/HDFS-6407?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Haohui Mai updated HDFS-6407: - Attachment: HDFS-6407.010.patch new namenode UI, lost ability to sort columns in datanode tab - Key: HDFS-6407 URL: https://issues.apache.org/jira/browse/HDFS-6407 Project: Hadoop HDFS Issue Type: Bug Components: namenode Affects Versions: 2.4.0 Reporter: Nathan Roberts Assignee: Haohui Mai Priority: Critical Labels: BB2015-05-TBR Attachments: 002-datanodes-sorted-capacityUsed.png, 002-datanodes.png, 002-filebrowser.png, 002-snapshots.png, HDFS-6407-002.patch, HDFS-6407-003.patch, HDFS-6407.008.patch, HDFS-6407.009.patch, HDFS-6407.010.patch, HDFS-6407.4.patch, HDFS-6407.5.patch, HDFS-6407.6.patch, HDFS-6407.7.patch, HDFS-6407.patch, browse_directory.png, datanodes.png, snapshots.png, sorting 2.png, sorting table.png old ui supported clicking on column header to sort on that column. The new ui seems to have dropped this very useful feature. There are a few tables in the Namenode UI to display datanodes information, directory listings and snapshots. When there are many items in the tables, it is useful to have ability to sort on the different columns. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648646#comment-14648646 ] Haohui Mai commented on HDFS-8833: -- bq. Currently we are copying repl from file level to block level.(HDFS-8823) I won't say this is a copy. The main reason why the replication factor is still there is that the {{FileStatus}} struct needs to report the replication factor. The BM only relies on information on the {{BlockInfo}} to make decisions on replication. I don't think that the EC branch will need to have the same compatibility concern. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8821) Explain message Operation category X is not supported in state standby
[ https://issues.apache.org/jira/browse/HDFS-8821?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648708#comment-14648708 ] Hudson commented on HDFS-8821: -- SUCCESS: Integrated in Hadoop-trunk-Commit #8247 (See [https://builds.apache.org/job/Hadoop-trunk-Commit/8247/]) HDFS-8821. Explain message Operation category X is not supported in state standby. Contributed by Gautam Gopalakrishnan. (harsh: rev c5caa25b8f2953e2b7a9d2c9dcbdbf1fed95c10b) * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/namenode/ha/StandbyState.java Explain message Operation category X is not supported in state standby - Key: HDFS-8821 URL: https://issues.apache.org/jira/browse/HDFS-8821 Project: Hadoop HDFS Issue Type: Improvement Reporter: Gautam Gopalakrishnan Assignee: Gautam Gopalakrishnan Priority: Minor Fix For: 2.8.0 Attachments: HDFS-8821-1.patch, HDFS-8821-2.patch There is one message specifically that causes many users to question the health of their HDFS cluster, namely Operation category READ/WRITE is not supported in state standby. HDFS-3447 is an attempt to lower the logging severity for StandbyException related messages but it is not resolved yet. So this jira is an attempt to explain this particular message so it appears less scary. The text is question 3.17 in the Hadoop Wiki FAQ ref: https://wiki.apache.org/hadoop/FAQ#What_does_the_message_.22Operation_category_READ.2FWRITE_is_not_supported_in_state_standby.22_mean.3F -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8059) Erasure coding: revisit how to store EC schema and cellSize in NameNode
[ https://issues.apache.org/jira/browse/HDFS-8059?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648644#comment-14648644 ] Haohui Mai commented on HDFS-8059: -- Sorry for the late reply. bq. Well, we keep replication in the namespace and use it in the block layer; how is this any different? I'm surprised you think that way. There are long efforts on separating the block manager out of the NN due to scalability concerns (which can be dated back since HDFS-2106). While we're not there yet but it is harmful to make new design choices that are contracted with the basic principle. bq. I assume you will solve the replication issue somehow, and the same solution should work for EC schema. Just quickly skim through the patch in this jira I think the current solution looks reasonable. For what needs to be done in trunk I'm putting up a patch in HDFS-8823. Erasure coding: revisit how to store EC schema and cellSize in NameNode --- Key: HDFS-8059 URL: https://issues.apache.org/jira/browse/HDFS-8059 Project: Hadoop HDFS Issue Type: Sub-task Affects Versions: HDFS-7285 Reporter: Yi Liu Assignee: Yi Liu Attachments: HDFS-8059.001.patch Move {{dataBlockNum}} and {{parityBlockNum}} from BlockInfoStriped to INodeFile, and store them in {{FileWithStripedBlocksFeature}}. Ideally these two nums are the same for all striped blocks in a file, and store them in BlockInfoStriped will waste NN memory. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-6407) new namenode UI, lost ability to sort columns in datanode tab
[ https://issues.apache.org/jira/browse/HDFS-6407?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648789#comment-14648789 ] Hadoop QA commented on HDFS-6407: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem || Runtime || Comment || | {color:blue}0{color} | pre-patch | 15m 29s | Pre-patch trunk compilation is healthy. | | {color:green}+1{color} | @author | 0m 0s | The patch does not contain any @author tags. | | {color:red}-1{color} | tests included | 0m 0s | 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. | | {color:green}+1{color} | javac | 7m 46s | There were no new javac warning messages. | | {color:green}+1{color} | javadoc | 9m 50s | There were no new javadoc warning messages. | | {color:green}+1{color} | release audit | 0m 22s | The applied patch does not increase the total number of release audit warnings. | | {color:green}+1{color} | whitespace | 0m 0s | The patch has no lines that end in whitespace. | | {color:green}+1{color} | install | 1m 21s | mvn install still works. | | {color:green}+1{color} | eclipse:eclipse | 0m 34s | The patch built with eclipse:eclipse. | | {color:green}+1{color} | native | 3m 3s | Pre-build of native portion | | {color:red}-1{color} | hdfs tests | 66m 52s | Tests failed in hadoop-hdfs. | | | | 105m 20s | | \\ \\ || Reason || Tests || | Failed unit tests | hadoop.hdfs.web.TestWebHdfsFileSystemContract | | Timed out tests | org.apache.hadoop.hdfs.TestHDFSFileSystemContract | \\ \\ || Subsystem || Report/Notes || | Patch URL | http://issues.apache.org/jira/secure/attachment/12748116/HDFS-6407.010.patch | | Optional Tests | javadoc javac unit | | git revision | trunk / c5caa25 | | hadoop-hdfs test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11876/artifact/patchprocess/testrun_hadoop-hdfs.txt | | Test Results | https://builds.apache.org/job/PreCommit-HDFS-Build/11876/testReport/ | | Java | 1.7.0_55 | | uname | Linux asf904.gq1.ygridcore.net 3.13.0-36-lowlatency #63-Ubuntu SMP PREEMPT Wed Sep 3 21:56:12 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux | | Console output | https://builds.apache.org/job/PreCommit-HDFS-Build/11876/console | This message was automatically generated. new namenode UI, lost ability to sort columns in datanode tab - Key: HDFS-6407 URL: https://issues.apache.org/jira/browse/HDFS-6407 Project: Hadoop HDFS Issue Type: Bug Components: namenode Affects Versions: 2.4.0 Reporter: Nathan Roberts Assignee: Haohui Mai Priority: Critical Labels: BB2015-05-TBR Attachments: 002-datanodes-sorted-capacityUsed.png, 002-datanodes.png, 002-filebrowser.png, 002-snapshots.png, HDFS-6407-002.patch, HDFS-6407-003.patch, HDFS-6407.008.patch, HDFS-6407.009.patch, HDFS-6407.010.patch, HDFS-6407.4.patch, HDFS-6407.5.patch, HDFS-6407.6.patch, HDFS-6407.7.patch, HDFS-6407.patch, browse_directory.png, datanodes.png, snapshots.png, sorting 2.png, sorting table.png old ui supported clicking on column header to sort on that column. The new ui seems to have dropped this very useful feature. There are a few tables in the Namenode UI to display datanodes information, directory listings and snapshots. When there are many items in the tables, it is useful to have ability to sort on the different columns. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8828) Utilize Snapshot diff report to build copy list in distcp
[ https://issues.apache.org/jira/browse/HDFS-8828?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Yufei Gu updated HDFS-8828: --- Description: Some users reported huge time cost to build file copy list in distcp. (30 hours for 1.6M files). We can leverage snapshot diff report to build file copy list including files/dirs which are changes only between two snapshots (or a snapshot and a normal dir). It speed up the process in two folds: 1. less copy list building time. 2. less file copy MR jobs. HDFS snapshot diff report provide information about file/directory creation, deletion, rename and modification between two snapshots or a snapshot and a normal directory. HDFS-7535 synchronize deletion and rename, then fallback to the default distcp. So it still relies on default distcp to building complete list of files under the source dir. This patch only puts creation and modification files into the copy list based on snapshot diff report. We can minimize the number of files to copy. was: Some users reported huge time cost to build file copy list in distcp. (30 hours with 1.6M files). We can leverage snapshot diff report to build file copy list including files/dirs which are changes only between two snapshots (or a snapshot and a normal dir). It speed up the process in two folds: 1. less copy list building time. 2. less file copy MR jobs. HDFS snapshot diff report provide information about file/directory creation, deletion, rename and modification between two snapshots or a snapshot and a normal directory. HDFS-7535 synchronize deletion and rename, then fallback to the default distcp. So it still relies on default distcp to building copy list which will traverse all files under the source dir. This patch will build the copy list based on snapshot diff report. Utilize Snapshot diff report to build copy list in distcp - Key: HDFS-8828 URL: https://issues.apache.org/jira/browse/HDFS-8828 Project: Hadoop HDFS Issue Type: Improvement Components: distcp, snapshots Reporter: Yufei Gu Assignee: Yufei Gu Attachments: HDFS-8828.001.patch Some users reported huge time cost to build file copy list in distcp. (30 hours for 1.6M files). We can leverage snapshot diff report to build file copy list including files/dirs which are changes only between two snapshots (or a snapshot and a normal dir). It speed up the process in two folds: 1. less copy list building time. 2. less file copy MR jobs. HDFS snapshot diff report provide information about file/directory creation, deletion, rename and modification between two snapshots or a snapshot and a normal directory. HDFS-7535 synchronize deletion and rename, then fallback to the default distcp. So it still relies on default distcp to building complete list of files under the source dir. This patch only puts creation and modification files into the copy list based on snapshot diff report. We can minimize the number of files to copy. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8830) Support add/remove directories to an existing encryption zone
[ https://issues.apache.org/jira/browse/HDFS-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xiaoyu Yao updated HDFS-8830: - Attachment: HDFS-8830.05.patch Delta from v4 patch: fix issue around rename added root dirs of encryption zone with additional unit tests added. Support add/remove directories to an existing encryption zone - Key: HDFS-8830 URL: https://issues.apache.org/jira/browse/HDFS-8830 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8830.01.patch, HDFS-8830.02.patch, HDFS-8830.03.patch, HDFS-8830.04.patch, HDFS-8830.05.patch This is the first step toward better Scratch space and Soft Delete support. We remove the assumption that the hdfs directory and encryption zone is 1 to 1 mapped and can't be changed once created. The encryption zone creation part is kept As-Is from Hadoop 2.4. We generalize the encryption zone and its directories from 1:1 to 1:many. This way, other directories such as scratch can be added to/removed from encryption zone as needed. Later on, files in these directories can be renamed within the same encryption zone efficiently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8830) Support add/remove directories to an existing encryption zone
[ https://issues.apache.org/jira/browse/HDFS-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647353#comment-14647353 ] Hadoop QA commented on HDFS-8830: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem || Runtime || Comment || | {color:blue}0{color} | pre-patch | 22m 37s | Pre-patch trunk compilation is healthy. | | {color:green}+1{color} | @author | 0m 0s | The patch does not contain any @author tags. | | {color:green}+1{color} | tests included | 0m 0s | The patch appears to include 2 new or modified test files. | | {color:green}+1{color} | javac | 8m 16s | There were no new javac warning messages. | | {color:green}+1{color} | javadoc | 10m 2s | There were no new javadoc warning messages. | | {color:green}+1{color} | release audit | 0m 24s | The applied patch does not increase the total number of release audit warnings. | | {color:green}+1{color} | checkstyle | 3m 50s | There were no new checkstyle issues. | | {color:green}+1{color} | whitespace | 0m 29s | The patch has no lines that end in whitespace. | | {color:green}+1{color} | install | 1m 31s | mvn install still works. | | {color:green}+1{color} | eclipse:eclipse | 0m 35s | The patch built with eclipse:eclipse. | | {color:red}-1{color} | findbugs | 4m 9s | Post-patch findbugs hadoop-hdfs-project/hadoop-hdfs compilation is broken. | | {color:red}-1{color} | findbugs | 4m 35s | Post-patch findbugs hadoop-hdfs-project/hadoop-hdfs-client compilation is broken. | | {color:green}+1{color} | findbugs | 4m 35s | The patch does not introduce any new Findbugs (version ) warnings. | | {color:red}-1{color} | common tests | 0m 25s | Tests failed in hadoop-common. | | {color:red}-1{color} | hdfs tests | 0m 27s | Tests failed in hadoop-hdfs. | | {color:green}+1{color} | hdfs tests | 0m 26s | Tests passed in hadoop-hdfs-client. | | | | 53m 41s | | \\ \\ || Reason || Tests || | Failed build | hadoop-common | | | hadoop-hdfs | \\ \\ || Subsystem || Report/Notes || | Patch URL | http://issues.apache.org/jira/secure/attachment/12747938/HDFS-8830.05.patch | | Optional Tests | javadoc javac unit findbugs checkstyle | | git revision | trunk / ddc867ce | | hadoop-common test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11873/artifact/patchprocess/testrun_hadoop-common.txt | | hadoop-hdfs test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11873/artifact/patchprocess/testrun_hadoop-hdfs.txt | | hadoop-hdfs-client test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11873/artifact/patchprocess/testrun_hadoop-hdfs-client.txt | | Test Results | https://builds.apache.org/job/PreCommit-HDFS-Build/11873/testReport/ | | Java | 1.7.0_55 | | uname | Linux asf904.gq1.ygridcore.net 3.13.0-36-lowlatency #63-Ubuntu SMP PREEMPT Wed Sep 3 21:56:12 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux | | Console output | https://builds.apache.org/job/PreCommit-HDFS-Build/11873/console | This message was automatically generated. Support add/remove directories to an existing encryption zone - Key: HDFS-8830 URL: https://issues.apache.org/jira/browse/HDFS-8830 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8830.01.patch, HDFS-8830.02.patch, HDFS-8830.03.patch, HDFS-8830.04.patch, HDFS-8830.05.patch This is the first step toward better Scratch space and Soft Delete support. We remove the assumption that the hdfs directory and encryption zone is 1 to 1 mapped and can't be changed once created. The encryption zone creation part is kept As-Is from Hadoop 2.4. We generalize the encryption zone and its directories from 1:1 to 1:many. This way, other directories such as scratch can be added to/removed from encryption zone as needed. Later on, files in these directories can be renamed within the same encryption zone efficiently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8792) Use LightWeightHashSet for BlockManager#postponedMisreplicatedBlocks
[ https://issues.apache.org/jira/browse/HDFS-8792?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Yi Liu updated HDFS-8792: - Attachment: HDFS-8792.002.patch Update a new patch. The new patch also improves {{excessReplicateMap}}, which we don't need {{sort}}, so no need to use a {{TreeMap}}. {code} public final MapString, LightWeightLinkedSetBlockInfo excessReplicateMap = -new TreeMap(); +new HashMap(); {code} Use LightWeightHashSet for BlockManager#postponedMisreplicatedBlocks Key: HDFS-8792 URL: https://issues.apache.org/jira/browse/HDFS-8792 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Yi Liu Assignee: Yi Liu Attachments: HDFS-8792.001.patch, HDFS-8792.002.patch {{LightWeightHashSet}} requires fewer memory than java hashset. Furthermore, for {{excessReplicateMap}}, we can use {{HashMap}} instead of {{TreeMap}} instead, since no need to sort. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647301#comment-14647301 ] Walter Su commented on HDFS-8833: - I see. You want to store ECPolicy at {{file level}} and also save a copy at {{zone/dir level}}. ( Or store at zone and save a copy at file) Currently StoragePolicy does the same thing. It save a copy at {{zone/dir level}} to ease the file creation. So we don't change the file creation API level. To me, StoragePolicy also has {{zone}} concept. It's just that storage zone/dir has loose limitation.(You can move a {{WARM}} file into a {{COLD}} directory.) ECPolicy just like StoragePolicy. A bit difference is: You can change the StoragePolicy of a directory/file any time you want. You can't change the ECPolicy of a file any time you want.(directory is fine, only affects file creation.) The title eliminate EC zones maybe not correct. As I said before, Three places to store schema and cellSize: {{zone level}}, {{file level}}, {{block level}}. StoragePolicy exists at {{zone level}}, {{file level}}. Currently we are copying repl from {{file level}} to {{block level}}.(HDFS-8823) Is it ok to store ECPolicy at {{zone level}}, {{file level}}, {{block level}} them all? I think it's ok. Becasue 1. {{zone level}} consumes little memory. 2. {{file level}} doesn't depends on {{zone level}}. If a zone is gone, the file still has a copy of the Policy. 3. namespace and storage layer should both have a copy. So two copies at least. 4. We give Policy an ID, so 3 copies consume little memory. I agree with [~zhz] that we store a copy of ECPolicy at {{file level}} like StoragePolicy did, for now. And in the future we can figure out how to copy StoragePolicy, ECPolicy, Repl together from namespace to storage layer. (HDFS-8823 has some progress) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8830) Support add/remove directories to an existing encryption zone
[ https://issues.apache.org/jira/browse/HDFS-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xiaoyu Yao updated HDFS-8830: - Attachment: (was: HDFS-8830.05.patch) Support add/remove directories to an existing encryption zone - Key: HDFS-8830 URL: https://issues.apache.org/jira/browse/HDFS-8830 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8830.01.patch, HDFS-8830.02.patch, HDFS-8830.03.patch, HDFS-8830.04.patch, HDFS-8830.05.patch This is the first step toward better Scratch space and Soft Delete support. We remove the assumption that the hdfs directory and encryption zone is 1 to 1 mapped and can't be changed once created. The encryption zone creation part is kept As-Is from Hadoop 2.4. We generalize the encryption zone and its directories from 1:1 to 1:many. This way, other directories such as scratch can be added to/removed from encryption zone as needed. Later on, files in these directories can be renamed within the same encryption zone efficiently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8830) Support add/remove directories to an existing encryption zone
[ https://issues.apache.org/jira/browse/HDFS-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xiaoyu Yao updated HDFS-8830: - Attachment: HDFS-8830.05.patch Support add/remove directories to an existing encryption zone - Key: HDFS-8830 URL: https://issues.apache.org/jira/browse/HDFS-8830 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8830.01.patch, HDFS-8830.02.patch, HDFS-8830.03.patch, HDFS-8830.04.patch, HDFS-8830.05.patch This is the first step toward better Scratch space and Soft Delete support. We remove the assumption that the hdfs directory and encryption zone is 1 to 1 mapped and can't be changed once created. The encryption zone creation part is kept As-Is from Hadoop 2.4. We generalize the encryption zone and its directories from 1:1 to 1:many. This way, other directories such as scratch can be added to/removed from encryption zone as needed. Later on, files in these directories can be renamed within the same encryption zone efficiently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (HDFS-488) Implement moveToLocal HDFS command
[ https://issues.apache.org/jira/browse/HDFS-488?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Steven Capo reassigned HDFS-488: Assignee: Steven Capo (was: Ravi Phulari) Implement moveToLocal HDFS command --- Key: HDFS-488 URL: https://issues.apache.org/jira/browse/HDFS-488 Project: Hadoop HDFS Issue Type: Bug Reporter: Ravi Phulari Assignee: Steven Capo Labels: newbie Attachments: Screen Shot 2014-07-23 at 12.28.23 PM 1.png Surprisingly executing HDFS FsShell command -moveToLocal outputs - Option '-moveToLocal' is not implemented yet. {code} statepick-lm:Hadoop rphulari$ bin/hadoop fs -moveToLocal bt t Option '-moveToLocal' is not implemented yet. {code} -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8792) Use LightWeightHashSet for BlockManager#postponedMisreplicatedBlocks
[ https://issues.apache.org/jira/browse/HDFS-8792?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647349#comment-14647349 ] Yi Liu commented on HDFS-8792: -- [~arpitagarwal], sorry for late response {quote} do you have any estimates of the memory saved by using LightWeightHashSet? {quote} Yes, compared to java {{HashSet}}, there are two advantages from memory point of review: # Java {{HashSet}} internally uses a {{HashMap}}, so there is one more reference (4 bytes) for each entry compared to {{LightWeightHashSet}}, so we can save {{4 * size}} bytes of memory. # In {{LightWeightHashSet}}, when elements become less, the size is shrinked a lot. So we can see {{LightWeightHashSet}} is more better. The main issue is {{LightWeightHashSet#LinkedSetIterator}} doesn't support {{remove}} currently, it's easy to support it (similar to java HashSet).By the way, currently in Hadoop, we use {{LightWeightHashSet}} for all big objects required hash set except this one which needs to use {{remove}}. Use LightWeightHashSet for BlockManager#postponedMisreplicatedBlocks Key: HDFS-8792 URL: https://issues.apache.org/jira/browse/HDFS-8792 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Yi Liu Assignee: Yi Liu Attachments: HDFS-8792.001.patch {{LightWeightHashSet}} requires fewer memory than java hashset. Furthermore, for {{excessReplicateMap}}, we can use {{HashMap}} instead of {{TreeMap}} instead, since no need to sort. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8792) Improve BlockManager#postponedMisreplicatedBlocks and BlockManager#excessReplicateMap
[ https://issues.apache.org/jira/browse/HDFS-8792?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Yi Liu updated HDFS-8792: - Summary: Improve BlockManager#postponedMisreplicatedBlocks and BlockManager#excessReplicateMap (was: Use LightWeightHashSet for BlockManager#postponedMisreplicatedBlocks) Improve BlockManager#postponedMisreplicatedBlocks and BlockManager#excessReplicateMap - Key: HDFS-8792 URL: https://issues.apache.org/jira/browse/HDFS-8792 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Yi Liu Assignee: Yi Liu Attachments: HDFS-8792.001.patch, HDFS-8792.002.patch {{LightWeightHashSet}} requires fewer memory than java hashset. Furthermore, for {{excessReplicateMap}}, we can use {{HashMap}} instead of {{TreeMap}} instead, since no need to sort. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648107#comment-14648107 ] Andrew Wang commented on HDFS-8833: --- bq. StoragePolicy do not have such restriction since moving a WARM file into a COLD dir as well changes its underlining storage (mover will move the replicas in order to satisfy the policy. IIUC this only happens if the file is set to inherit and picked up WARM from a parent directory. If the storage policy was set on the file itself, it would not change when renamed under a COLD directory. The proposal is basically like storage policies, except there is no inherit mode. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648073#comment-14648073 ] Tsz Wo Nicholas Sze commented on HDFS-8833: --- To me, StoragePolicy also has zone concept. It's just that storage zone/dir has loose limitation.(You can move a WARM file into a COLD directory.) ... When talking about zone, we mean that files in one zone cannot be (easily) moved/chnaged to another zone. StoragePolicy do not have such restriction since moving a WARM file into a COLD dir as well changes its underlining storage (mover will move the replicas in order to satisfy the policy.) However, we current cannot change the file encryption scheme and the erasure code scheme after a file is created. So we need zones for them. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8480) Fix performance and timeout issues in HDFS-7929 by using hard-links to preserve old edit logs instead of copying them
[ https://issues.apache.org/jira/browse/HDFS-8480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648076#comment-14648076 ] Colin Patrick McCabe commented on HDFS-8480: Good catch, [~mingma]. If we make a change later to the edit log format, this test will break because we'll be writing out an edit log with the new format, but the old version ID. I think we should simply check in a small binary edit log in order to do this testing. This is similar to what we do with other ugprade scenario tests. As [~zhz] mentioned, it would be a lot of effort to support writing out edit logs in older formats (we have literally dozens of versions). Fix performance and timeout issues in HDFS-7929 by using hard-links to preserve old edit logs instead of copying them - Key: HDFS-8480 URL: https://issues.apache.org/jira/browse/HDFS-8480 Project: Hadoop HDFS Issue Type: Bug Affects Versions: 2.7.0 Reporter: Zhe Zhang Assignee: Zhe Zhang Priority: Critical Fix For: 2.7.1 Attachments: HDFS-8480.00.patch, HDFS-8480.01.patch, HDFS-8480.02.patch, HDFS-8480.03.patch HDFS-7929 copies existing edit logs to the storage directory of the upgraded {{NameNode}}. This slows down the upgrade process. This JIRA aims to use hard-linking instead of per-op copying to achieve the same goal. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648084#comment-14648084 ] Tsz Wo Nicholas Sze commented on HDFS-8833: --- What is the semantic of moving a file under EC zone A to EC zone B? Would the file be changed from EC scheme A to EC schema B? If yes, we could eliminate EC zones. Otherwise, we should keep EC zone. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8747) Provide Better Scratch Space and Soft Delete Support for HDFS Encryption Zones
[ https://issues.apache.org/jira/browse/HDFS-8747?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648086#comment-14648086 ] Andrew Wang commented on HDFS-8747: --- Hi [~xyao], I like this idea overall, nice work. A few questions: * Have you thought about simply allowing rename between EZs with the same settings? This would be a much smaller and easier change with similar properties. Your proposal I think is still better in terms of ease-of-use and also ensuring security invariants around key rolling (if/when we implement that). * If we keep the APIs superuser-only, how does a normal user add their trash folder to an EZ? Same for scratch folders, e.g. if the Hive user is not a superuser. Provide Better Scratch Space and Soft Delete Support for HDFS Encryption Zones -- Key: HDFS-8747 URL: https://issues.apache.org/jira/browse/HDFS-8747 Project: Hadoop HDFS Issue Type: Bug Components: encryption Affects Versions: 2.6.0 Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8747-07092015.pdf, HDFS-8747-07152015.pdf, HDFS-8747-07292015.pdf HDFS Transparent Data Encryption At-Rest was introduced in Hadoop 2.6 to allow create encryption zone on top of a single HDFS directory. Files under the root directory of the encryption zone will be encrypted/decrypted transparently upon HDFS client write or read operations. Generally, it does not support rename(without data copying) across encryption zones or between encryption zone and non-encryption zone because different security settings of encryption zones. However, there are certain use cases where efficient rename support is desired. This JIRA is to propose better support of two such use cases “Scratch Space” (a.k.a. staging area) and “Soft Delete” (a.k.a. trash) with HDFS encryption zones. “Scratch Space” is widely used in Hadoop jobs, which requires efficient rename support. Temporary files from MR jobs are usually stored in staging area outside encryption zone such as “/tmp” directory and then rename to targeted directories as specified once the data is ready to be further processed. Below is a summary of supported/unsupported cases from latest Hadoop: * Rename within the encryption zone is supported * Rename the entire encryption zone by moving the root directory of the zone is allowed. * Rename sub-directory/file from encryption zone to non-encryption zone is not allowed. * Rename sub-directory/file from encryption zone A to encryption zone B is not allowed. * Rename from non-encryption zone to encryption zone is not allowed. “Soft delete” (a.k.a. trash) is a client-side “soft delete” feature that helps prevent accidental deletion of files and directories. If trash is enabled and a file or directory is deleted using the Hadoop shell, the file is moved to the .Trash directory of the user's home directory instead of being deleted. Deleted files are initially moved (renamed) to the Current sub-directory of the .Trash directory with original path being preserved. Files and directories in the trash can be restored simply by moving them to a location outside the .Trash directory. Due to the limited rename support, delete sub-directory/file within encryption zone with trash feature is not allowed. Client has to use -skipTrash option to work around this. HADOOP-10902 and HDFS-6767 improved the error message but without a complete solution to the problem. We propose to solve the problem by generalizing the mapping between encryption zone and its underlying HDFS directories from 1:1 today to 1:N. The encryption zone should allow non-overlapped directories such as scratch space or soft delete trash locations to be added/removed dynamically after creation. This way, rename for scratch space and soft delete can be better supported without breaking the assumption that rename is only supported within the zone. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648088#comment-14648088 ] Tsz Wo Nicholas Sze commented on HDFS-8833: --- Anyway, we don't seem to have a complete design yet for eliminating EC zones. If we really want to do it, we should discuss the design first before changing any code. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8838) Tolerate datanode failures in DFSStripedOutputStream when the data length is small
[ https://issues.apache.org/jira/browse/HDFS-8838?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648095#comment-14648095 ] Tsz Wo Nicholas Sze commented on HDFS-8838: --- Hi [~libo-intel], HDFS-8704 seems to fix large files according to the JIRA summary but here is to fix small files. Am I missing anything? Tolerate datanode failures in DFSStripedOutputStream when the data length is small -- Key: HDFS-8838 URL: https://issues.apache.org/jira/browse/HDFS-8838 Project: Hadoop HDFS Issue Type: Sub-task Components: hdfs-client Reporter: Tsz Wo Nicholas Sze Assignee: Tsz Wo Nicholas Sze Attachments: h8838_20150729.patch Currently, DFSStripedOutputStream cannot tolerate datanode failures when the data length is small. We fix the bugs here and add more tests. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8704) Erasure Coding: client fails to write large file when one datanode fails
[ https://issues.apache.org/jira/browse/HDFS-8704?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648099#comment-14648099 ] Tsz Wo Nicholas Sze commented on HDFS-8704: --- +for (int dn = 0; dn 1; dn++) { Please try killing all the datanodes but not just one. Otherwise, we cannot completely test the fix. Erasure Coding: client fails to write large file when one datanode fails Key: HDFS-8704 URL: https://issues.apache.org/jira/browse/HDFS-8704 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Li Bo Assignee: Li Bo Attachments: HDFS-8704-000.patch, HDFS-8704-HDFS-7285-002.patch, HDFS-8704-HDFS-7285-003.patch I test current code on a 5-node cluster using RS(3,2). When a datanode is corrupt, client succeeds to write a file smaller than a block group but fails to write a large one. {{TestDFSStripeOutputStreamWithFailure}} only tests files smaller than a block group, this jira will add more test situations. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8704) Erasure Coding: client fails to write large file when one datanode fails
[ https://issues.apache.org/jira/browse/HDFS-8704?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647239#comment-14647239 ] Li Bo commented on HDFS-8704: - I have just update a second patch of this problem. The changes in this patch include: 1. {{DFSStripedOutputStream}} and the failed status of {{StripedDataStreamer}}: it’s not right to take different actions according to the current status of a streamer. When a streamer has failed, while the packet to be queued belongs to the next block(the streamer can successfully write to that block because the new datanode may be well), in this condition the packet should be handled as usual. When {{DFSStripedOutputStream}} finds a streamer is working well, it queue the packet to the streamer, but the streamer may fail before sending the packet. So I remove the logic of checking and setting the failed status of a streamer in {{DFSStripedOutputStream}}. When a streamer fails, itself knows how to handle the failure. 2. Extend the functionality of {{StripedDataStreamer}} : if error occurs, {{ StripedDataStreamer }} will first handle remaining trivial packets of current block, and then restart to waiting for a new block to be allocated to it. 3. Add a test to {{TestDFSStripedOutputStreamWithFailure}} which tests writing a file with two block groups. The unit test occasionally fails because only 8 block locations are given by namenode for the second block group. HDFS-8839 has been created to track this problem. Erasure Coding: client fails to write large file when one datanode fails Key: HDFS-8704 URL: https://issues.apache.org/jira/browse/HDFS-8704 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Li Bo Assignee: Li Bo Attachments: HDFS-8704-000.patch, HDFS-8704-HDFS-7285-002.patch I test current code on a 5-node cluster using RS(3,2). When a datanode is corrupt, client succeeds to write a file smaller than a block group but fails to write a large one. {{TestDFSStripeOutputStreamWithFailure}} only tests files smaller than a block group, this jira will add more test situations. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647251#comment-14647251 ] Zhe Zhang commented on HDFS-8833: - {code} hdfs erasurecode [generic options] [-setECPolicy [-s schemaName] [-c cellSize] path] [-getECPolicy path] [-listSchemas] [-usage [cmd ...]] [-help [cmd ...]] SetECPolicy command: -setECPolicy [-s schemaName] [-c cellSize] path SetECPolicy command is used to set an erasure coding policy, including codec schema and cell size, at the specified path. path: Refer to an empty and already created directory in HDFS. This is a mandatory parameter. schemaName: This is an optional parameter, specified using ‘-s’ flag. Refer to the name of ECSchema to be used for erasure coding of direct children (files or subdirectories) of this directory. If not specified the system default ECSchema will be used. cellSize: This is also an optional parameter, specified using ‘-c’ flag. Refer to the size of cells for striped blocks. If not specified the default cellsize (64k) will be used. GetECPolicy command [-getECPolicy path] GetECPolicy command is used to get details of the erasure coding policy, including codec schema and cell size, for the specified path ListSchemas command: [-listSchemas] Lists all ECSchemas supported for erasure coding. For SetECPolicy command, one of these ECSchema’s name should be provided. {code} Above is the proposed CLI. I took the current version and made a few modifications. ([~vinayrpet] could you take a look and see if I'm missing anything? Thanks..) From admin's perspective the only difference is nested policies (renaming is a result of it). Renaming is just to deviate from the existing expectations on _zones_ established by encryption zones. From an implementation perspective the main difference is to store a copy of inherited EC policy in the XAttr of {{INodeFile}}. As Walter reiterated above about listing ECPolicies, I think it's very likely that we can limit the number of supported polices to 16~256 so we can fit the policy ID (4~8 bits) in file header. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-6697) Make NN lease soft and hard limits configurable
[ https://issues.apache.org/jira/browse/HDFS-6697?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647262#comment-14647262 ] Vinayakumar B commented on HDFS-6697: - bq. Can you clarify why in LeaseRenewer it needs to compare the configured soft limit with the default soft limit? Instead it can just set the renewal variable with the configured value. I think the idea is to keep {{renawal}} interval to max 30 seconds(default). And if configured softlimit is positive and less than 60 seconds, that will have the priority. Am I right [~andreina]? Make NN lease soft and hard limits configurable --- Key: HDFS-6697 URL: https://issues.apache.org/jira/browse/HDFS-6697 Project: Hadoop HDFS Issue Type: Improvement Reporter: Ming Ma Assignee: J.Andreina Attachments: HDFS-6697.1.patch, HDFS-6697.2.patch, HDFS-6697.3.patch For testing, NameNodeAdapter allows test code to specify lease soft and hard limit via setLeasePeriod directly on LeaseManager. But NamenodeProxies.java still use the default values. It is useful if we can make NN lease soft and hard limit configurable via Configuration. That will allow NamenodeProxies.java to use the configured values. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8835) Convert BlockInfoUnderConstruction as an interface
[ https://issues.apache.org/jira/browse/HDFS-8835?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647269#comment-14647269 ] Zhe Zhang commented on HDFS-8835: - Thanks Nicholas for the suggestion. This JIRA shares the same purpose with HDFS-8487. In HDFS-7285 we made a lot of changes to generalize {{BlockInfo}}, which are not directly related to EC. Pushing those changes to trunk separately will make ease merge reviewing (especially for reviewers not directly involved in branch dev). So I'll hold on working on a patch, and leave this JIRA open to solicit feedbacks from merge reviewing. LMK your thoughts. Thanks. Convert BlockInfoUnderConstruction as an interface -- Key: HDFS-8835 URL: https://issues.apache.org/jira/browse/HDFS-8835 Project: Hadoop HDFS Issue Type: Improvement Components: namenode Affects Versions: 2.7.1 Reporter: Zhe Zhang Assignee: Zhe Zhang Per discussion under HDFS-8499, this JIRA aims to convert {{BlockInfoUnderConstruction}} as an interface and {{BlockInfoContiguousUnderConstruction}} as its implementation. The HDFS-7285 branch will add {{BlockInfoStripedUnderConstruction}} as another implementation. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8698) Add -direct flag option for fs copy so that user can choose not to create ._COPYING_ file
[ https://issues.apache.org/jira/browse/HDFS-8698?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Vinayakumar B updated HDFS-8698: Issue Type: Improvement (was: New Feature) Add -direct flag option for fs copy so that user can choose not to create ._COPYING_ file - Key: HDFS-8698 URL: https://issues.apache.org/jira/browse/HDFS-8698 Project: Hadoop HDFS Issue Type: Improvement Components: fs Affects Versions: 2.7.0 Reporter: Chen He Assignee: J.Andreina Because CLI is using CommandWithDestination.java which add ._COPYING_ to the tail of file name when it does the copy. For blobstore like S3 and Swift, to create ._COPYING_ file and rename it is expensive. -direct flag can allow user to avoiding the ._COPYING_ file. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647996#comment-14647996 ] Andrew Wang commented on HDFS-8833: --- Great discussion [~zhz] [~walter.k.su] and [~vinayrpet], agree with everything so far. In terms of memory overhead, as long as it only increases memory usage of EC files, I'm okay with it. One hybrid approach we could pursue is providing a set of hardcoded policies and then allow additional user-configured policies as an XAttr. Users that stick to the the hardcoded policies will have improved memory usage. We can also dedupe xattr values in addition to names as an additional optimization. Since IIUC we only want to provide hardcoded policies in phase 1, this hybrid approach can come later. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8823) Move replication factor into individual blocks
[ https://issues.apache.org/jira/browse/HDFS-8823?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14648137#comment-14648137 ] Haohui Mai commented on HDFS-8823: -- bq. Thanks for the work Haohui. As Walter Su pointed out , the patch copies rather than moves the replication factor. So a minor ask is to update the JIRA description. Yes there is duplicate information but I don't feel this is a copy since the BM only relies on the information the BlockInfo for replication? I plan to further clean things up in subsequent patches. bq. To give my two cents: replication factor, storage policy, and EC policy (potentially) are the same class of control knobs for HDFS users and we should probably consolidate the discussions on how to manage them in the context of separating NN and BM layers. There are long efforts and consensus that BM layer and the namespace need to be separated (e.g., HDFS-2106, HDFS-5477 and HDFS-7836). Though progress is slow but we should not make design and implementation choices that contradicts with the effort. You might want to take a look of the related jiras and the code to get more contexts. This is a nontrivial effort and IMO it is more constructive to discuss in separate jiras with concrete contexts. Speaking of what you've mentioned, this particular jira proposes to decouple BM and NN w.r.t replication factors in trunk. I've given my feedbacks in the HDFS-8833 already. There is no jira w.r.t. the storage policy yet. However, you're more than welcomed to take a crack and submit a patch if you want to investigate on storage policy. Move replication factor into individual blocks -- Key: HDFS-8823 URL: https://issues.apache.org/jira/browse/HDFS-8823 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Attachments: HDFS-8823.000.patch This jira proposes to record the replication factor in the {{BlockInfo}} class. The changes have two advantages: * Decoupling the namespace and the block management layer. It is a prerequisite step to move block management off the heap or to a separate process. * Increased flexibility on replicating blocks. Currently the replication factors of all blocks have to be the same. The replication factors of these blocks are equal to the highest replication factor across all snapshots. The changes will allow blocks in a file to have different replication factor, potentially saving some space. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8816) Improve visualization for the Datanode tab in the NN UI
[ https://issues.apache.org/jira/browse/HDFS-8816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647448#comment-14647448 ] Hudson commented on HDFS-8816: -- FAILURE: Integrated in Hadoop-Yarn-trunk #1002 (See [https://builds.apache.org/job/Hadoop-Yarn-trunk/1002/]) HDFS-8816. Improve visualization for the Datanode tab in the NN UI. Contributed by Haohui Mai. (wheat9: rev ddc867ceb9a76986e8379361753598cc48024376) * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.html * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/hadoop.css * hadoop-hdfs-project/hadoop-hdfs/CHANGES.txt * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/hdfs/dfshealth.js * hadoop-hdfs-project/hadoop-hdfs/pom.xml * hadoop-hdfs-project/hadoop-hdfs/src/main/webapps/static/moment.min.js Improve visualization for the Datanode tab in the NN UI --- Key: HDFS-8816 URL: https://issues.apache.org/jira/browse/HDFS-8816 Project: Hadoop HDFS Issue Type: Improvement Reporter: Haohui Mai Assignee: Haohui Mai Fix For: 2.8.0 Attachments: HDFS-8816.000.patch, HDFS-8816.001.patch, HDFS-8816.002.patch, HDFS-8816.003.patch, HDFS-8816.004.patch, HDFS-8816.png, HDFS-8816.png, Screen Shot 2015-07-23 at 10.24.24 AM.png The information of the datanode tab in the NN UI is clogged. This jira proposes to improve the visualization of the datanode tab in the UI. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8388) Time and Date format need to be in sync in Namenode UI page
[ https://issues.apache.org/jira/browse/HDFS-8388?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647452#comment-14647452 ] Surendra Singh Lilhore commented on HDFS-8388: -- Thanks [~ajisakaa] for review.. bq. 2. Would you use ddd MMM DD HH:mm:ss ZZ format instead of ddd MMM DD HH:mm:ss to be consistent with YARN Web UI? If I use ddd MMM DD HH:mm:ss ZZ format then Block Deletion Start Time and Compiled time is coming like this {noformat} Compiled: Mon Jul 06 12:20:00 +0530 2015 by sslilhore from trunk Block Deletion Start Time: Thu Jul 30 14:31:30 +0530 2015 {noformat} and Started time is like this {noformat} Started:Thu Jul 30 17:01:30 CST 2015 {noformat} here Compiled time zone is +0530 but Started time zone is CST. is this ok ?? Time and Date format need to be in sync in Namenode UI page --- Key: HDFS-8388 URL: https://issues.apache.org/jira/browse/HDFS-8388 Project: Hadoop HDFS Issue Type: Bug Reporter: Archana T Assignee: Surendra Singh Lilhore Priority: Minor Attachments: HDFS-8388-002.patch, HDFS-8388.patch, HDFS-8388_1.patch, ScreenShot-InvalidDate.png In NameNode UI Page, Date and Time FORMAT displayed on the page are not in sync currently. Started:Wed May 13 12:28:02 IST 2015 Compiled:23 Apr 2015 12:22:59 Block Deletion Start Time 13 May 2015 12:28:02 We can keep a common format in all the above places. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8830) Support add/remove directories to an existing encryption zone
[ https://issues.apache.org/jira/browse/HDFS-8830?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647464#comment-14647464 ] Hadoop QA commented on HDFS-8830: - \\ \\ | (x) *{color:red}-1 overall{color}* | \\ \\ || Vote || Subsystem || Runtime || Comment || | {color:blue}0{color} | pre-patch | 20m 49s | Pre-patch trunk compilation is healthy. | | {color:green}+1{color} | @author | 0m 0s | The patch does not contain any @author tags. | | {color:green}+1{color} | tests included | 0m 0s | The patch appears to include 2 new or modified test files. | | {color:green}+1{color} | javac | 7m 35s | There were no new javac warning messages. | | {color:green}+1{color} | javadoc | 9m 38s | There were no new javadoc warning messages. | | {color:green}+1{color} | release audit | 0m 23s | The applied patch does not increase the total number of release audit warnings. | | {color:green}+1{color} | checkstyle | 3m 27s | There were no new checkstyle issues. | | {color:green}+1{color} | whitespace | 0m 26s | The patch has no lines that end in whitespace. | | {color:green}+1{color} | install | 1m 37s | mvn install still works. | | {color:green}+1{color} | eclipse:eclipse | 0m 33s | The patch built with eclipse:eclipse. | | {color:red}-1{color} | findbugs | 6m 17s | The patch appears to introduce 1 new Findbugs (version 3.0.0) warnings. | | {color:green}+1{color} | common tests | 22m 14s | Tests passed in hadoop-common. | | {color:red}-1{color} | hdfs tests | 162m 6s | Tests failed in hadoop-hdfs. | | {color:green}+1{color} | hdfs tests | 0m 27s | Tests passed in hadoop-hdfs-client. | | | | 235m 37s | | \\ \\ || Reason || Tests || | FindBugs | module:hadoop-hdfs | | Failed unit tests | hadoop.hdfs.TestRollingUpgrade | \\ \\ || Subsystem || Report/Notes || | Patch URL | http://issues.apache.org/jira/secure/attachment/12747933/HDFS-8830.05.patch | | Optional Tests | javadoc javac unit findbugs checkstyle | | git revision | trunk / ddc867ce | | Findbugs warnings | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/artifact/patchprocess/newPatchFindbugsWarningshadoop-hdfs.html | | hadoop-common test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/artifact/patchprocess/testrun_hadoop-common.txt | | hadoop-hdfs test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/artifact/patchprocess/testrun_hadoop-hdfs.txt | | hadoop-hdfs-client test log | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/artifact/patchprocess/testrun_hadoop-hdfs-client.txt | | Test Results | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/testReport/ | | Java | 1.7.0_55 | | uname | Linux asf909.gq1.ygridcore.net 3.13.0-36-lowlatency #63-Ubuntu SMP PREEMPT Wed Sep 3 21:56:12 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux | | Console output | https://builds.apache.org/job/PreCommit-HDFS-Build/11872/console | This message was automatically generated. Support add/remove directories to an existing encryption zone - Key: HDFS-8830 URL: https://issues.apache.org/jira/browse/HDFS-8830 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Reporter: Xiaoyu Yao Assignee: Xiaoyu Yao Attachments: HDFS-8830.01.patch, HDFS-8830.02.patch, HDFS-8830.03.patch, HDFS-8830.04.patch, HDFS-8830.05.patch This is the first step toward better Scratch space and Soft Delete support. We remove the assumption that the hdfs directory and encryption zone is 1 to 1 mapped and can't be changed once created. The encryption zone creation part is kept As-Is from Hadoop 2.4. We generalize the encryption zone and its directories from 1:1 to 1:many. This way, other directories such as scratch can be added to/removed from encryption zone as needed. Later on, files in these directories can be renamed within the same encryption zone efficiently. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8838) Tolerate datanode failures in DFSStripedOutputStream when the data length is small
[ https://issues.apache.org/jira/browse/HDFS-8838?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647366#comment-14647366 ] Li Bo commented on HDFS-8838: - Thanks Nicholas for the effort on this problem. I have just updated a new patch to HDFS-8704 which also fixes this problem. I will review the patch of this jira and hope we can merge them together. Tolerate datanode failures in DFSStripedOutputStream when the data length is small -- Key: HDFS-8838 URL: https://issues.apache.org/jira/browse/HDFS-8838 Project: Hadoop HDFS Issue Type: Sub-task Components: hdfs-client Reporter: Tsz Wo Nicholas Sze Assignee: Tsz Wo Nicholas Sze Attachments: h8838_20150729.patch Currently, DFSStripedOutputStream cannot tolerate datanode failures when the data length is small. We fix the bugs here and add more tests. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647360#comment-14647360 ] Vinayakumar B commented on HDFS-8833: - {quote}path: Refer to an empty and already created directory in HDFS. This is a mandatory parameter. schemaName: This is an optional parameter, specified using ‘-s’ flag. Refer to the name of ECSchema to be used for erasure coding of direct children (files or subdirectories) of this directory. If not specified the system default ECSchema will be used.{quote} Here, description for path and schemaName should be updated. It could be file also. And need not be empty. I think can change after creating files also, only thing is further files will inherit different ECPolicy. Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (HDFS-8704) Erasure Coding: client fails to write large file when one datanode fails
[ https://issues.apache.org/jira/browse/HDFS-8704?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Li Bo updated HDFS-8704: Attachment: HDFS-8704-HDFS-7285-003.patch Erasure Coding: client fails to write large file when one datanode fails Key: HDFS-8704 URL: https://issues.apache.org/jira/browse/HDFS-8704 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Li Bo Assignee: Li Bo Attachments: HDFS-8704-000.patch, HDFS-8704-HDFS-7285-002.patch, HDFS-8704-HDFS-7285-003.patch I test current code on a 5-node cluster using RS(3,2). When a datanode is corrupt, client succeeds to write a file smaller than a block group but fails to write a large one. {{TestDFSStripeOutputStreamWithFailure}} only tests files smaller than a block group, this jira will add more test situations. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8704) Erasure Coding: client fails to write large file when one datanode fails
[ https://issues.apache.org/jira/browse/HDFS-8704?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647363#comment-14647363 ] Li Bo commented on HDFS-8704: - Patch 003 fixes the problem of writing a file that is a little smaller than a block group, (ie. file length = blocksize * dataBlocks -123). If the test suspends, that's mainly caused by HDFS-8839. Erasure Coding: client fails to write large file when one datanode fails Key: HDFS-8704 URL: https://issues.apache.org/jira/browse/HDFS-8704 Project: Hadoop HDFS Issue Type: Sub-task Reporter: Li Bo Assignee: Li Bo Attachments: HDFS-8704-000.patch, HDFS-8704-HDFS-7285-002.patch, HDFS-8704-HDFS-7285-003.patch I test current code on a 5-node cluster using RS(3,2). When a datanode is corrupt, client succeeds to write a file smaller than a block group but fails to write a large one. {{TestDFSStripeOutputStreamWithFailure}} only tests files smaller than a block group, this jira will add more test situations. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (HDFS-8833) Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones
[ https://issues.apache.org/jira/browse/HDFS-8833?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=14647371#comment-14647371 ] Vinayakumar B commented on HDFS-8833: - Idea of having ECPolicy is good. But copying the Same content in All levels, (dir, file and block), will significantly increase the memory usage of the NameNode. I think this was one of the motivation for erasure coding zone concept. I agree that, trash and rename have some problems. As [~walter.k.su] mentioned, we can have a ID, to save the memery usage and use this ID everywhere, as done for BlockStoragePolicy. But BlockStoragePolicy is limited to max 16 policies and its hardcoded. I dont think we can hardcode all possible schemas and limit to only 16. If its made configurable, IDs of the already existing schemas should not change at all. Huge increase in NameNode memory usage is by default not accepted. [~zhz], you have any possible estimation, how much memory usage will be affected from this change? Erasure coding: store EC schema and cell size with INodeFile and eliminate EC zones --- Key: HDFS-8833 URL: https://issues.apache.org/jira/browse/HDFS-8833 Project: Hadoop HDFS Issue Type: Sub-task Components: namenode Affects Versions: HDFS-7285 Reporter: Zhe Zhang Assignee: Zhe Zhang We have [discussed | https://issues.apache.org/jira/browse/HDFS-7285?focusedCommentId=14357754page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14357754] storing EC schema with files instead of EC zones and recently revisited the discussion under HDFS-8059. As a recap, the _zone_ concept has severe limitations including renaming and nested configuration. Those limitations are valid in encryption for security reasons and it doesn't make sense to carry them over in EC. This JIRA aims to store EC schema and cell size on {{INodeFile}} level. For simplicity, we should first implement it as an xattr and consider memory optimizations (such as moving it to file header) as a follow-on. We should also disable changing EC policy on a non-empty file / dir in the first phase. -- This message was sent by Atlassian JIRA (v6.3.4#6332)