[
https://issues.apache.org/jira/browse/HBASE-866?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12672717#action_12672717
]
Andrew Purtell commented on HBASE-866:
--------------------------------------
Ok, I have figured out what is going on, at least in my case.
OOME in IPC server handler causes the IPC handler to abort, but the client
never learns about this, so it waits and waits and waits... I have seen
Heritrix writer threads that have been waiting for 7+ hours. And, the OOME does
not take down the HRS, so it stays up in some degraded state. See HBASE-1196.
Sorry to pollute this issue.
> Blocking for ten minutes at a time
> ----------------------------------
>
> Key: HBASE-866
> URL: https://issues.apache.org/jira/browse/HBASE-866
> Project: Hadoop HBase
> Issue Type: Bug
> Reporter: stack
> Priority: Minor
> Attachments: thread_dump.txt
>
>
> I've been testing running biggish MR jobs uploading into hbase. My jobs
> consistently fail with child task timing out its ten minute period. Adding
> logging, was able to see that we're actual stuck in a commit. Following the
> thread of the row we're committing, I see this in the log:
> {code}
> ...
> 2008-09-03 18:37:03,446 DEBUG org.apache.hadoop.hbase.regionserver.HRegion:
> Flush requested on TestTable,0029377106,1220466998108
> 2008-09-03 18:37:03,446 DEBUG org.apache.hadoop.hbase.regionserver.HRegion:
> Started memcache flush for region TestTable,0029377106,1220466998108. Current
> region memcache size 64.0m
> 2008-09-03 18:37:03,446 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 1 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:13,450 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 1 on 60020'
> 2008-09-03 18:37:16,089 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 16 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 1 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 4 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 6 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 2 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 12 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,090 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 9 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:16,091 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Blocking updates for 'IPC Server handler 7 on 60020' on region
> TestTable,0029377106,1220466998108: Memcache size 64.0m is >= than blocking
> 64.0m size
> 2008-09-03 18:37:21,984 DEBUG org.apache.hadoop.hbase.regionserver.HRegion:
> Finished memcache flush for region TestTable,0029377106,1220466998108 in
> 18538ms, sequence id=2852547, compaction requested=false
> 2008-09-03 18:47:06,241 DEBUG org.apache.hadoop.hbase.regionserver.HRegion:
> Started memcache flush for region TestTable,0029377106,1220466998108. Current
> region memcache size 64.0m
> 2008-09-03 18:47:10,031 DEBUG org.apache.hadoop.hbase.regionserver.HRegion:
> Finished memcache flush for region TestTable,0029377106,1220466998108 in
> 3790ms, sequence id=2919208, compaction requested=true
> 2008-09-03 18:47:10,031 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 9 on 60020'
> 2008-09-03 18:47:10,031 DEBUG
> org.apache.hadoop.hbase.regionserver.CompactSplitThread: Compaction requested
> for region: TestTable,0029377106,1220466998108
> 2008-09-03 18:47:10,031 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 12 on 60020'
> 2008-09-03 18:47:10,032 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> starting compaction on region TestTable,0029377106,1220466998108
> 2008-09-03 18:47:10,032 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 7 on 60020'
> 2008-09-03 18:47:10,035 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 6 on 60020'
> 2008-09-03 18:47:10,035 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 4 on 60020'
> 2008-09-03 18:47:10,035 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 2 on 60020'
> 2008-09-03 18:47:10,037 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 16 on 60020'
> 2008-09-03 18:47:10,043 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> Unblocking updates for region TestTable,0029377106,1220466998108 'IPC Server
> handler 1 on 60020'
> 2008-09-03 18:47:18,403 INFO org.apache.hadoop.hbase.regionserver.HRegion:
> compaction completed on region TestTable,0029377106,1220466998108 in 8sec
> ...
> {code}
> Notice how we're blocked for ten minutes until new flush runs. My guess is
> that the flush that is going on concurrent with the blocking is clearing the
> flag
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
