[ https://issues.apache.org/jira/browse/HBASE-15265?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15216662#comment-15216662 ]
stack commented on HBASE-15265: ------------------------------- It took me a while but a 2B run on 9-node cluster passes with this enabled (on Master -- I got Master to pass ITBLL... in this case Master is not carrying regions... not sure that related). Will try some more runs. I see this stuff in the log: 511740 2016-03-29 10:00:05,239 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 483 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-c70edb05-4a60-4d80-a8b7-5783f707a9da,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-6539429a-1441-45e8-9d47-6732499f4709,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-b3225e5d-2c3a-4cf3-a91d-10b345451c24,DISK]] 511741 2016-03-29 10:00:05,239 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 482 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-c70edb05-4a60-4d80-a8b7-5783f707a9da,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-6539429a-1441-45e8-9d47-6732499f4709,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-b3225e5d-2c3a-4cf3-a91d-10b345451c24,DISK]] 511742 2016-03-29 10:00:05,239 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 475 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-c70edb05-4a60-4d80-a8b7-5783f707a9da,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-6539429a-1441-45e8-9d47-6732499f4709,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-b3225e5d-2c3a-4cf3-a91d-10b345451c24,DISK]] Seems like long time to sync but I see these w/o Async enabled... so just HDFS... to be investigated. Or these... {code} 511541 2016-03-29 09:59:54,779 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 761 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-59aa4068-698e-4c2d-8f37-d921d8cbd3ec,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-42e19728-db67-4b51-9cd2-10a3b640a8e5,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-188eee65-528f-4ced-84ac-894d634590f7,DISK]] 511542 2016-03-29 09:59:54,779 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 760 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-59aa4068-698e-4c2d-8f37-d921d8cbd3ec,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-42e19728-db67-4b51-9cd2-10a3b640a8e5,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-188eee65-528f-4ced-84ac-894d634590f7,DISK]] 511543 2016-03-29 09:59:54,779 INFO [AsyncFSWAL-pool3-t1] wal.AbstractFSWAL: Slow sync cost: 760 ms, current pipeline: [DatanodeInfoWithStorage[10.17.240.22:50031,DS-59aa4068-698e-4c2d-8f37-d921d8cbd3ec,DISK], DatanodeInfoWithStorage[10.17.240.27:50031,DS-42e19728-db67-4b51-9cd2-10a3b640a8e5,DISK], DatanodeInfoWithStorage[10.17.240.25:50031,DS-188eee65-528f-4ced-84ac-894d634590f7,DISK]] {code} ... need to dig in on the above. File issue to make this the default? Meantime I'll run some more tests. Nice work [~Apache9] > Implement an asynchronous FSHLog > -------------------------------- > > Key: HBASE-15265 > URL: https://issues.apache.org/jira/browse/HBASE-15265 > Project: HBase > Issue Type: Sub-task > Components: wal > Reporter: Duo Zhang > Assignee: Duo Zhang > Fix For: 2.0.0 > > Attachments: HBASE-15265-v1.patch, HBASE-15265-v2.patch, > HBASE-15265-v3.patch, HBASE-15265-v4.patch, HBASE-15265-v5.patch, > HBASE-15265-v6.patch, HBASE-15265-v7.patch, HBASE-15265-v8.patch, > HBASE-15265-v8.patch, HBASE-15265-v8.patch, HBASE-15265-v8.patch, > HBASE-15265-v8.patch, HBASE-15265-v8.patch, HBASE-15265-v8.patch, > HBASE-15265-v8.patch, HBASE-15265.patch > > -- This message was sent by Atlassian JIRA (v6.3.4#6332)