[
https://issues.apache.org/jira/browse/HBASE-2646?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12915952#action_12915952
]
Ted Yu commented on HBASE-2646:
-------------------------------
Summarizing discussion with Jeff:
Ted: In PriorityCompactionQueue.addToRegionsInQueue(), I noticed the following
call which is not synchronized:
queue.remove(queuedRequest);
Now suppose before the above is executed, PriorityCompactionQueue.take() is
called. So queuedRequest is returned to the caller of take(). Later, this line
in take():
removeFromRegionsInQueue(cr.getHRegion());
would remove the newly added, higher priority request from regionsInQueue.
Jeff:
That is an astute observation. Stepping through the code with the threads
stopping execution at the points in code you suggest would indeed make it so
take() would return the lower priority compactionRequest, remove the higher
priority compaction request from regionsInQueue, and finally the add() method
would complete and add the higher priority compaction onto the queue with no
corresponding entry in the regionsInQueue hash (this is bad). Even if I move
the queue.remove(queuedRequest) into the synchronized(regionsInQueue) we will
run into the same problem.
Fortunately the worst thing that can happen is there is a request that doesn't
have an entry in regionsInQueue that will eventually be executed with no
adverse problem for the system other than extra work. This wont actually cause
any problems to the system but PriorityCompactionQueue will have an
inconsistent state which should be fixed. An immediate solution is not jumping
out at me. So I need to think through the problem and see if I can't come up
with a way to prevent the inconsistency.
Ted:
Except for remove(Object r), all callers of removeFromRegionsInQueue() have
CompactionRequest information.
So CompactionRequest, cr, can be passed to removeFromRegionsInQueue() so that
we can perform some sanity check.
If cr.getPriority() is lower than the priority of the CompactionRequest
currently in regionsInQueue, removeFromRegionsInQueue() can return null which
indicates inconsistency.
The caller can discard cr upon seeing null from removeFromRegionsInQueue() and
try to get the next request from queue.
The above avoids introducing another synchronization between accesses to queue
and regionsInQueue.
Jeff:
I was thinking along the same lines. Adding an additional synchronization
didn't seem like the right approach. So if we make sure we are taking off what
we are expecting to then there wont be a problem.
> Compaction requests should be prioritized to prevent blocking
> -------------------------------------------------------------
>
> Key: HBASE-2646
> URL: https://issues.apache.org/jira/browse/HBASE-2646
> Project: HBase
> Issue Type: Improvement
> Components: regionserver
> Affects Versions: 0.20.4
> Environment: ubuntu server 10; hbase 0.20.4; 4 machine cluster (each
> machine is an 8 core xeon with 16 GB of ram and 6TB of storage); ~250 Million
> rows;
> Reporter: Jeff Whiting
> Priority: Critical
> Fix For: 0.20.7
>
> Attachments: 2646-v2.txt, 2646-v3.txt,
> prioritycompactionqueue-0.20.4.patch, PriorityQueue-r996664.patch
>
>
> While testing the write capacity of a 4 machine hbase cluster we were getting
> long and frequent client pauses as we attempted to load the data. Looking
> into the problem we'd get a relatively large compaction queue and when a
> region hit the "hbase.hstore.blockingStoreFiles" limit it would get block the
> client and the compaction request would get put on the back of the queue
> waiting for many other less important compactions. The client is basically
> stuck at that point until a compaction is done. Prioritizing the compaction
> requests and allowing the request that is blocking other actions go first
> would help solve the problem.
> You can see the problem by looking at our log files:
> You'll first see an event such as a too many HLog which will put a lot of
> requests on the compaction queue.
> {noformat}
> 2010-05-25 10:53:26,570 INFO org.apache.hadoop.hbase.regionserver.HLog: Too
> many hlogs: logs=33, maxlogs=32; forcing flush of 22 regions(s):
> responseCounts,RS_6eZzLtdwhGiTwHy,1274232223324,
> responses,RS_0qhkL5rUmPCbx3K-1274213057242,1274513189592,
> responses,RS_1ANYnTegjzVIsHW-12742177419
> 21,1274511001873, responses,RS_1HQ4UG5BdOlAyuE-1274216757425,1274726323747,
> responses,RS_1Y7SbqSTsZrYe7a-1274328697838,1274478031930,
> responses,RS_1ZH5TB5OdW4BVLm-1274216239894,1274538267659,
> responses,RS_3BHc4KyoM3q72Yc-1274290546987,1274502062319,
> responses,RS_3ra9BaBMAXFAvbK-127421457
> 9958,1274381552543, responses,RS_6SDrGNuyyLd3oR6-1274219941155,1274385453586,
> responses,RS_8AGCEMWbI6mZuoQ-1274306857429,1274319602718,
> responses,RS_8C8T9DN47uwTG1S-1274215381765,1274289112817,
> responses,RS_8J5wmdmKmJXzK6g-1274299593861,1274494738952,
> responses,RS_8e5Sz0HeFPAdb6c-1274288
> 641459,1274495868557,
> responses,RS_8rjcnmBXPKzI896-1274306981684,1274403047940,
> responses,RS_9FS3VedcyrF0KX2-1274245971331,1274754745013,
> responses,RS_9oZgPtxO31npv3C-1274214027769,1274396489756,
> responses,RS_a3FdO2jhqWuy37C-1274209228660,1274399508186,
> responses,RS_a3LJVxwTj29MHVa-12742
> {noformat}
> Then you see the too many log files:
> {noformat}
> 2010-05-25 10:53:31,364 DEBUG
> org.apache.hadoop.hbase.regionserver.CompactSplitThread: Compaction requested
> for region
> responses-index,--1274799047787--R_cBKrGxx0FdWjPso,1274804575862/783020138
> because: regionserver/192.168.0.81:60020.cacheFlusher
> 2010-05-25 10:53:32,364 WARN
> org.apache.hadoop.hbase.regionserver.MemStoreFlusher: Region
> responses-index,--1274799047787--R_cBKrGxx0FdWjPso,1274804575862 has too many
> store files, putting it back at the end of the flush queue.
> {noformat}
> Which leads to this:
> {noformat}
> 2010-05-25 10:53:27,061 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 60 on 60020' on region
> responses-index,--1274799047787--R_cBKrGxx0FdWjPso,1274804575862: memstore
> size 128.0m is >= than blocking 128.0m size
> 2010-05-25 10:53:27,061 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 84 on 60020' on region
> responses-index,--1274799047787--R_cBKrGxx0FdWjPso,1274804575862: memstore
> size 128.0m is >= than blocking 128.0m size
> 2010-05-25 10:53:27,065 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 1 on 60020' on region
> responses-index,--1274799047787--R_cBKrGxx0FdWjPso,1274804575862: memstore
> size 128.0m is >= than blocking 128.0m size
> {noformat}
> Once the compaction / split is done a flush is able to happen which unblocks
> the IPC allowing writes to continue. Unfortunately this process can take
> upwards of 15+ minutes (the specific case shown here from our logs took about
> 4 minutes).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
