[jira] [Commented] (HBASE-22547) Align the config keys and add document for offheap read in HBase Book.

[ https://issues.apache.org/jira/browse/HBASE-22547?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16869223#comment-16869223 ] Zheng Hu commented on HBASE-22547: -- Pushed to HBASE-21879 branch, Thanks [~anoop.hbase] for his

[jira] [Updated] (HBASE-22612) Address the final overview reviewing comments of HBASE-21879

[ https://issues.apache.org/jira/browse/HBASE-22612?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22612: - Description: I've created a big PR(https://github.com/apache/hbase/pull/320) for HBASE-21879, and

[jira] [Created] (HBASE-22612) Address the final overview reviewing comments of HBASE-21879

Zheng Hu created HBASE-22612: Summary: Address the final overview reviewing comments of HBASE-21879 Key: HBASE-22612 URL: https://issues.apache.org/jira/browse/HBASE-22612 Project: HBase Issue

[jira] [Updated] (HBASE-22547) Align the config keys and add document for offheap read in HBase Book.

[ https://issues.apache.org/jira/browse/HBASE-22547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22547: - Attachment: HBASE-22547.HBASE-21879.v1.patch > Align the config keys and add document for offheap read

[jira] [Updated] (HBASE-22547) Align the config keys and add document for offheap read in HBase Book.

[ https://issues.apache.org/jira/browse/HBASE-22547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22547: - Status: Patch Available (was: Open) > Align the config keys and add document for offheap read in HBase

[jira] [Updated] (HBASE-22547) Align the config keys and add document for offheap read in HBase Book.

[ https://issues.apache.org/jira/browse/HBASE-22547?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22547: - Summary: Align the config keys and add document for offheap read in HBase Book. (was: Document for

[jira] [Updated] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22561: - Resolution: Fixed Hadoop Flags: Reviewed Status: Resolved (was: Patch Available) >

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16867309#comment-16867309 ] Zheng Hu commented on HBASE-22561: -- Pushed to all branches, Thanks [~dbist13] for contributing. >

[jira] [Updated] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22561: - Fix Version/s: 2.1.6 2.2.1 2.0.6 2.3.0

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16867267#comment-16867267 ] Zheng Hu commented on HBASE-22561: -- Patch.v12 looks good to me now, +1. Let me commit ( will fix the

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16866137#comment-16866137 ] Zheng Hu commented on HBASE-22561: -- For HFilePrettyPrinter, seems no need to add the -u option, also

[jira] [Resolved] (HBASE-22598) Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility

[ https://issues.apache.org/jira/browse/HBASE-22598?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22598. -- Resolution: Fixed Hadoop Flags: Reviewed > Deprecated the

[jira] [Commented] (HBASE-22598) Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility

[ https://issues.apache.org/jira/browse/HBASE-22598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865600#comment-16865600 ] Zheng Hu commented on HBASE-22598: -- Pushed to HBASE-21879 branch, Thanks [~reidchan] & [~anoop.hbase]

[jira] [Commented] (HBASE-22598) Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility

[ https://issues.apache.org/jira/browse/HBASE-22598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865485#comment-16865485 ] Zheng Hu commented on HBASE-22598: -- Thanks for your comment about the LOG.WARN suggestion. Created a

[jira] [Comment Edited] (HBASE-22598) Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility

[ https://issues.apache.org/jira/browse/HBASE-22598?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865485#comment-16865485 ] Zheng Hu edited comment on HBASE-22598 at 6/17/19 9:57 AM: --- Thanks for your

[jira] [Created] (HBASE-22598) Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility

Zheng Hu created HBASE-22598: Summary: Deprecated the hbase.ipc.server.reservoir.initial.buffer.size & hbase.ipc.server.reservoir.initial.max for HBase2.x compatibility Key: HBASE-22598 URL:

[jira] [Commented] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865418#comment-16865418 ] Zheng Hu commented on HBASE-22582: -- OK, assigned to myself, will upload patch for this bug. > The

[jira] [Assigned] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu reassigned HBASE-22582: Assignee: Zheng Hu > The Compaction writer may access the lastCell whose memory has been

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865391#comment-16865391 ] Zheng Hu commented on HBASE-22561: -- The patch.v7 Looks good to me now. It may be better if moving

[jira] [Resolved] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22387. -- Resolution: Fixed > Evaluate the get/scan performance after reading HFile block into offheap >

[jira] [Commented] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16865283#comment-16865283 ] Zheng Hu commented on HBASE-22387: -- I've updated all performances result in the desgin doc, please see

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16863987#comment-16863987 ] Zheng Hu commented on HBASE-22561: -- FYI [~mackrorysd] & [~mantonov] > modify HFilePrettyPrinter to

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16863980#comment-16863980 ] Zheng Hu commented on HBASE-22561: -- Thanks for your patch, [~dbist13]. Seems the santiy check was

[jira] [Commented] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16863809#comment-16863809 ] Zheng Hu commented on HBASE-22582: -- [~javaman_chen], you plan to prepare a fix & UT for this(If sure

[jira] [Commented] (HBASE-21879) Read HFile's block to ByteBuffer directly instead of to byte for reducing young gc purpose

[ https://issues.apache.org/jira/browse/HBASE-21879?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16863797#comment-16863797 ] Zheng Hu commented on HBASE-21879: -- Nice finding, Thanks [~javaman_chen], It's indeed a bug, I've

[jira] [Updated] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22582: - Fix Version/s: 2.1.6 2.2.1 2.0.6 3.0.0 > The

[jira] [Updated] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22582: - Component/s: Compaction > The Compaction writer may access the lastCell whose memory has been released

[jira] [Updated] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

[ https://issues.apache.org/jira/browse/HBASE-22582?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22582: - Description: Copy the comment from [~javaman_chen] under HBASE-21879:

[jira] [Created] (HBASE-22582) The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final

Zheng Hu created HBASE-22582: Summary: The Compaction writer may access the lastCell whose memory has been released when appending fileInfo in the final Key: HBASE-22582 URL:

[jira] [Resolved] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22531. -- Resolution: Fixed Hadoop Flags: Reviewed > The HFileReaderImpl#shouldUseHeap return the

[jira] [Commented] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16863600#comment-16863600 ] Zheng Hu commented on HBASE-22531: -- Pushed to HBASE-21879 branch, Thanks all for reviewing. > The

[jira] [Commented] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16862768#comment-16862768 ] Zheng Hu commented on HBASE-22531: -- Let's see what HBaseQA says, if OK , I will commit this patch.

[jira] [Resolved] (HBASE-22491) Separate the heap HFileBlock and offheap HFileBlock because the heap block won't need refCnt and save into prevBlocks list before shipping

[ https://issues.apache.org/jira/browse/HBASE-22491?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22491. -- Resolution: Fixed Hadoop Flags: Reviewed Pushed to branch HBASE-21879, Thanks [~anoop.hbase] &

[jira] [Commented] (HBASE-21879) Read HFile's block to ByteBuffer directly instead of to byte for reducing young gc purpose

[ https://issues.apache.org/jira/browse/HBASE-21879?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16861699#comment-16861699 ] Zheng Hu commented on HBASE-21879: -- [~elserj], This feature is almost done now, I think it will be

[jira] [Commented] (HBASE-22561) modify HFilePrettyPrinter to accept non-hbase.rootdir directories

[ https://issues.apache.org/jira/browse/HBASE-22561?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16861678#comment-16861678 ] Zheng Hu commented on HBASE-22561: -- Mind you prepare a patch for this ? [~dbist13]. > modify

[jira] [Commented] (HBASE-22559) [RPC] set guard against CALL_QUEUE_HANDLER_FACTOR_CONF_KEY

[ https://issues.apache.org/jira/browse/HBASE-22559?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16861662#comment-16861662 ] Zheng Hu commented on HBASE-22559: -- Nice finding, maybe we also need a check to ensure that the

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16857399#comment-16857399 ] Zheng Hu commented on HBASE-22532: -- bq. lets make sure others benefit from your findings sir. Yeah,

[jira] [Created] (HBASE-22547) Document for offheap read in HBase Book.

Zheng Hu created HBASE-22547: Summary: Document for offheap read in HBase Book. Key: HBASE-22547 URL: https://issues.apache.org/jira/browse/HBASE-22547 Project: HBase Issue Type: Sub-task

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16856782#comment-16856782 ] Zheng Hu commented on HBASE-22532: -- Addressed this issue in design doc[1] of HBASE-21879 today, Will

[jira] [Resolved] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22532. -- Resolution: Not A Problem > There's still too much cpu wasting on validating checksum even if >

[jira] [Commented] (HBASE-22539) Potential WAL corruption due to Unsafe.copyMemory usage when DBB are in place

[ https://issues.apache.org/jira/browse/HBASE-22539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16856277#comment-16856277 ] Zheng Hu commented on HBASE-22539: -- It's strange that ByteBufferUtils.copyFromBufferToArray would made

[jira] [Commented] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855549#comment-16855549 ] Zheng Hu commented on HBASE-22531: -- It's not the time now, because the current patch.v1 depends on

[jira] [Commented] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855535#comment-16855535 ] Zheng Hu commented on HBASE-22387: -- [~anoop.hbase], Yeah, filed two HDFS issues for this: HDFS-14535

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855470#comment-16855470 ] Zheng Hu commented on HBASE-22532: -- Yeah, the HBASE-21879 has already addressed your concern. Since

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855457#comment-16855457 ] Zheng Hu commented on HBASE-22532: -- Our non-data ( bloom_chunk or leaf_index ) block were > 128KB,

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855452#comment-16855452 ] Zheng Hu commented on HBASE-22532: -- There are some LEAF_INDEX block too: {code} $ cat out.log | grep

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855447#comment-16855447 ] Zheng Hu commented on HBASE-22532: -- OK, seems all of the block whose size > 65KB are BLOOM_CHUNK data

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Attachment: async-prof-pid-64695-cpu-1.svg > There's still too much cpu wasting on validating checksum

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855298#comment-16855298 ] Zheng Hu commented on HBASE-22532: -- After I increased the hbase.ipc.server.allocator.buffer.size from

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855281#comment-16855281 ] Zheng Hu commented on HBASE-22532: -- Enabled the TRACE log level for package:

[jira] [Commented] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16855223#comment-16855223 ] Zheng Hu commented on HBASE-22531: -- bq. You meant off heap Zheng Hu? Sorry, I think I did not

[jira] [Comment Edited] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16854521#comment-16854521 ] Zheng Hu edited comment on HBASE-22531 at 6/4/19 1:45 AM: -- After applied the

[jira] [Commented] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16854604#comment-16854604 ] Zheng Hu commented on HBASE-22532: -- Add a log like the following:  {code} commit

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Description: After disabled the block cache, and with the following config: {code} # Disable the

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Description: After disabled the block cache, and with the following config: {code} # Disable the

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Description: After disabled the block cache, and with the following config: {code} # Disable the

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Attachment: async-prof-pid-27827-cpu-3.svg > There's still too much cpu wasting on validating checksum

[jira] [Updated] (HBASE-22532) There's still too much cpu wasting on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Summary: There's still too much cpu wasting on validating checksum even if buffer.size=65KB (was:

[jira] [Created] (HBASE-22532) There're still too much cpu cost on validating checksum even if buffer.size=65KB

Zheng Hu created HBASE-22532: Summary: There're still too much cpu cost on validating checksum even if buffer.size=65KB Key: HBASE-22532 URL: https://issues.apache.org/jira/browse/HBASE-22532 Project:

[jira] [Updated] (HBASE-22532) There's still too much cpu cost on validating checksum even if buffer.size=65KB

[ https://issues.apache.org/jira/browse/HBASE-22532?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22532: - Summary: There's still too much cpu cost on validating checksum even if buffer.size=65KB (was:

[jira] [Updated] (HBASE-21946) Replace the byte[] pread by ByteBuffer pread in HFileBlock reading once HDFS-3246 prepared

[ https://issues.apache.org/jira/browse/HBASE-21946?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-21946: - Attachment: HBASE-21946.HBASE-21879.v04.patch > Replace the byte[] pread by ByteBuffer pread in

[jira] [Comment Edited] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16854521#comment-16854521 ] Zheng Hu edited comment on HBASE-22531 at 6/3/19 12:15 PM: --- After applied the

[jira] [Commented] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16854521#comment-16854521 ] Zheng Hu commented on HBASE-22531: -- After applied the patch.v1, seems all of the allocation happened on

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Attachment: async-prof-pid-8590-alloc-2.svg > The HFileReaderImpl#shouldUseHeap return the incorrect

[jira] [Updated] (HBASE-21946) Replace the byte[] pread by ByteBuffer pread in HFileBlock reading once HDFS-3246 prepared

[ https://issues.apache.org/jira/browse/HBASE-21946?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-21946: - Attachment: HBASE-21946.HBASE-21879.v03.patch > Replace the byte[] pread by ByteBuffer pread in

[jira] [Updated] (HBASE-21946) Replace the byte[] pread by ByteBuffer pread in HFileBlock reading once HDFS-3246 prepared

[ https://issues.apache.org/jira/browse/HBASE-21946?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-21946: - Attachment: HBASE-21946.HBASE-21879.v02.patch > Replace the byte[] pread by ByteBuffer pread in

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Attachment: HBASE-22531.HBASE-21879.v1.patch > The HFileReaderImpl#shouldUseHeap return the incorrect

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Description: I'm having a benchmark with block cache disabled for HBASE-21879 branch. Just caurious

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Description: I'm having a benchmark with block cache disabled for HBASE-21879 branch. Just caurious

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Description: I'm having a benchmark with block cache disabled for HBASE-21879 branch. Just caurious

[jira] [Updated] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

[ https://issues.apache.org/jira/browse/HBASE-22531?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22531: - Attachment: async-prof-pid-13311-alloc-4.svg > The HFileReaderImpl#shouldUseHeap return the incorrect

[jira] [Created] (HBASE-22531) The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache

Zheng Hu created HBASE-22531: Summary: The HFileReaderImpl#shouldUseHeap return the incorrect true when disabled BlockCache Key: HBASE-22531 URL: https://issues.apache.org/jira/browse/HBASE-22531

[jira] [Comment Edited] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851639#comment-16851639 ] Zheng Hu edited comment on HBASE-22387 at 5/31/19 11:36 AM: Upload three

[jira] [Comment Edited] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851639#comment-16851639 ] Zheng Hu edited comment on HBASE-22387 at 5/31/19 11:35 AM: Upload three

[jira] [Updated] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22387: - Attachment: async-prof-pid-24534-alloc-2.svg > Evaluate the get/scan performance after reading HFile

[jira] [Resolved] (HBASE-22483) It's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22483. -- Resolution: Fixed Hadoop Flags: Reviewed > It's better to use 65KB as the default buffer size

[jira] [Commented] (HBASE-22483) It's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16852869#comment-16852869 ] Zheng Hu commented on HBASE-22483: -- Pushed to HBASE-21879 branch, thanks [~ram_krish] & [~Apache9] for

[jira] [Commented] (HBASE-22483) It's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16852798#comment-16852798 ] Zheng Hu commented on HBASE-22483: -- [~ram_krish], Yeah, Attached the PR just now, please take a look.

[jira] [Updated] (HBASE-22483) It's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22483: - Summary: It's better to use 65KB as the default buffer size in ByteBuffAllocator (was: Maybe it's

[jira] [Updated] (HBASE-22483) Maybe it's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22483: - Description: There're some reason why it's better to choose 65KB as the default buffer size: 1. Almost

[jira] [Commented] (HBASE-22483) Maybe it's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16852786#comment-16852786 ] Zheng Hu commented on HBASE-22483: -- Had two benchmark with hbase.ipc.server.allocator.buffer.size =

[jira] [Updated] (HBASE-22483) Maybe it's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22483: - Attachment: with-buffer-size-64KB.png > Maybe it's better to use 65KB as the default buffer size in

[jira] [Updated] (HBASE-22483) Maybe it's better to use 65KB as the default buffer size in ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22483?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22483: - Attachment: with-buffer-size-65KB.png > Maybe it's better to use 65KB as the default buffer size in

[jira] [Resolved] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[ https://issues.apache.org/jira/browse/HBASE-22504?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22504. -- Resolution: Fixed Hadoop Flags: Reviewed > Optimize the MultiByteBuff#get(ByteBuffer, offset,

[jira] [Commented] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[ https://issues.apache.org/jira/browse/HBASE-22504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16852701#comment-16852701 ] Zheng Hu commented on HBASE-22504: -- Pushed to HBASE-21879 branch, Thanks [~ram_krish] & [~anoop.hbase]

[jira] [Commented] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[ https://issues.apache.org/jira/browse/HBASE-22504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16852696#comment-16852696 ] Zheng Hu commented on HBASE-22504: -- Checked all API in MBB, seems other API are optimized now.  Let me

[jira] [Updated] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[ https://issues.apache.org/jira/browse/HBASE-22504?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22504: - Attachment: HBASE-22504.HBASE-21879.v01.patch > Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[jira] [Comment Edited] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851705#comment-16851705 ] Zheng Hu edited comment on HBASE-22387 at 5/30/19 1:48 PM: --- Some issues we've

[jira] [Updated] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22387: - Attachment: 0001-Initialize-a-small-buffer-to-send-the-short-circuit-.patch > Evaluate the get/scan

[jira] [Updated] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

[ https://issues.apache.org/jira/browse/HBASE-22504?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22504: - Description: In HBASE-22483,  we saw that the BucketCacheWriter thread was quite busy

[jira] [Created] (HBASE-22504) Optimize the MultiByteBuff#get(ByteBuffer, offset, len)

Zheng Hu created HBASE-22504: Summary: Optimize the MultiByteBuff#get(ByteBuffer, offset, len) Key: HBASE-22504 URL: https://issues.apache.org/jira/browse/HBASE-22504 Project: HBase Issue Type:

[jira] [Commented] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851705#comment-16851705 ] Zheng Hu commented on HBASE-22387: -- Some issues we've found in the flamegraph: 1. About 45% heap

[jira] [Updated] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22387: - Attachment: screenshot-2.png > Evaluate the get/scan performance after reading HFile block into offheap

[jira] [Comment Edited] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851639#comment-16851639 ] Zheng Hu edited comment on HBASE-22387 at 5/30/19 8:39 AM: --- Upload three

[jira] [Commented] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851639#comment-16851639 ] Zheng Hu commented on HBASE-22387: -- Upload three profile svg graphs with our HBASE-21879 branch

[jira] [Updated] (HBASE-22387) Evaluate the get/scan performance after reading HFile block into offheap directly

[ https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22387: - Attachment: async-prof-pid-25042-alloc-2.svg async-prof-pid-25042-cpu-1.svg

[jira] [Resolved] (HBASE-22463) Some paths in HFileScannerImpl did not consider block#release which will exhaust the ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22463?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu resolved HBASE-22463. -- Resolution: Fixed Hadoop Flags: Reviewed > Some paths in HFileScannerImpl did not consider

[jira] [Commented] (HBASE-22463) Some paths in HFileScannerImpl did not consider block#release which will exhaust the ByteBuffAllocator

[ https://issues.apache.org/jira/browse/HBASE-22463?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851596#comment-16851596 ] Zheng Hu commented on HBASE-22463: -- Pushed to branch HBASE-21879, Thanks [~anoop.hbase] & [~Apache9]

[jira] [Commented] (HBASE-22496) UnsafeAccess.unsafeCopy should not copy more than UNSAFE_COPY_THRESHOLD on each iteration

[ https://issues.apache.org/jira/browse/HBASE-22496?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16851472#comment-16851472 ] Zheng Hu commented on HBASE-22496: -- Pushed to all branches, Thanks [~wchevreuil] for contributing, and

[jira] [Updated] (HBASE-22496) UnsafeAccess.unsafeCopy should not copy more than UNSAFE_COPY_THRESHOLD on each iteration

[ https://issues.apache.org/jira/browse/HBASE-22496?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Zheng Hu updated HBASE-22496: - Resolution: Fixed Hadoop Flags: Reviewed Status: Resolved (was: Patch Available) >

<    1   2   3   4   5   6   7   8   9   10   >