Uh oh!
There was an error while loading. Please reload this page.
HBASE-25720 Sync WAL stuck when prepare flush cache will prevent flus… - #3140
HBASE-25720 Sync WAL stuck when prepare flush cache will prevent flus…#3140sunhelly wants to merge 3 commits into
Conversation
Apache-HBase
commented
Apr 9, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Apr 9, 2021
💔 -1 overall
This message was automatically generated. |
Apache-HBase
commented
Apr 9, 2021
💔 -1 overall
This message was automatically generated. |
Apache-HBase
commented
Apr 11, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Apr 11, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Apr 11, 2021
💔 -1 overall
This message was automatically generated. |
Apache-HBase
commented
May 17, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
May 17, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
May 17, 2021
🎊 +1 overall
This message was automatically generated. |
| doSyncOfUnflushedWALChanges(wal, getRegionInfo()); | ||
| } catch (Throwable t) { | ||
| status.abort("Sync unflushed WAL changes failed: " + StringUtils.stringifyException(t)); | ||
| fatalForFlushCache(t); |
There was a problem hiding this comment.
Mind explaining a bit why here we need to abort the regionserver? We do not need to abort the regionserver when hitting IOException above?
There was a problem hiding this comment.
The IOException thrown above is not caused by syncing WAL, only the sync failure should abort the RS as early as possible. And if we only allow aborting RS in the commit step of flushing cache, there will always exit the circumstance that OOM kill, because all failures in the previous preparing step of flushing cache only aborted the flush process, all flushing can not go to the commit step.
There was a problem hiding this comment.
@Apache9 We currently only abort RS when there are critical problems, such as the WAL stuck? The above IOExceptions can only abort the flush in mem and should not prevent the next flushes, right? Here the abort is caused by the wal stucks who are ahead of the committing flush wal sync stucks. What do you think?
Apache-HBase
commented
May 26, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
May 26, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
May 26, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Sep 16, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Sep 16, 2021
🎊 +1 overall
This message was automatically generated. |
Apache-HBase
commented
Sep 16, 2021
🎊 +1 overall
This message was automatically generated. |
…hing cache and cause OOM