[jira] [Updated] (HBASE-8665) bad compaction priority behavior in queue can cause store to be blocked

["Sergey Shelukhin (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HBASE-8665?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sergey Shelukhin updated HBASE-8665:
------------------------------------

    Attachment: HBASE-8665-v0.patch

Here's the patch. It's rather simple, most complexity is in tests.
It allows for both pre-selected compactions (from coprocessors and users) and non-pre-selected.
When non-preselected compaction makes it to run() it is selected and executed.
There are two special cases:
1) Store priority might have decreased since the compaction was queued; in that case it's re-queued with new priority to avoid inversion.
2) Without selecting, we don't know which pool to go to. We go to small pool by default, and if the compaction is large after selection, queue it to the large pool; it can bounce back if it becomes small while stuck in the large pool.

Both of these cases can cause compaction to get continuously requeued, or bounced between pools, however it can only happen if store priority decreases (i.e. other compactions happen), or files are removed from store (same); or files are added in some very special pattern that causes policy to select large-small-large-small-... continuously).
                
> bad compaction priority behavior in queue can cause store to be blocked
> -----------------------------------------------------------------------
>
>                 Key: HBASE-8665
>                 URL: https://issues.apache.org/jira/browse/HBASE-8665
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Sergey Shelukhin
>            Assignee: Sergey Shelukhin
>         Attachments: HBASE-8665-v0.patch
>
>
> Note that this can be solved by bumping up the number of compaction threads but still it seems like this priority "inversion" should be dealt with.
> There's a store with 1 big file and 3 flushes (1 2 3 4) sitting around and minding its own business when it decides to compact. Compaction (2 3 4) is created and put in queue, it's low priority, so it doesn't get out of the queue for some time - other stores are compacting. Meanwhile more files are flushed and at (1 2 3 4 5 6 7) it decides to compact (5 6 7). This compaction now has higher priority than the first one. After that if the load is high it enters vicious cycle of compacting and compacting files as they arrive, with store being blocked on and off, with the (2 3 4) compaction staying in queue for up to ~20 minutes (that I've seen).
> I wonder why we do thing thing where we queue compaction and compact separately. Perhaps we should take snapshot of all store priorities, then do select in order and execute the first compaction we find. This will need starvation safeguard too but should probably be better.
> Btw, exploring compaction policy may be more prone to this, as it can select files from the middle, not just beginning, which, given the treatment of already selected files that was not changed from the old ratio-based one (all files with lower seqNums than the ones selected are also ineligible for further selection), will make more files ineligible (e.g. imagine with 10 blocking files, with 8 present (1-8), (6 7 8) being selected and getting stuck). Today I see the case that would also apply to old policy, but yesterday I saw file distribution something like this: 4,5g, 2,1g, 295,9m, 113,3m, 68,0m, 67,8m, 1,1g, 295,1m, 100,4m, unfortunately w/o enough logs to figure out how it resulted.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira