> Thanks Brian, > is there any clue in logs spanning 2014-09-08 06:25 +/- 30 minutes?
I don’t see anything interesting at this point in the log files. Based on some application timeouts I was seeing, I think the problem began several days earlier. I looked in the area where these messages first began to appear, but nothing jumped out at me; a write-heavy MR job was running at the time, so there might be something buried in the noise (a lot of noise). > > > 在 2014-9-11，19:08，Brian Jeltema <brian.jelt...@digitalenvoy.net> 写道： > > the RS log is huge. What do you want to see other than what was posted > earlier? > > On Sep 11, 2014, at 6:10 AM, Qiang Tian <tian...@gmail.com> wrote: > > that is strange. we hit it in 0.94 as well: > > https://issues.apache.org/jira/browse/HBASE-11695 > > what Nicolas raised makes sense to me, but revisiting the code, the flush > > every 10s in the RS log actually comes from HRegion#shouldFlush, so there > > is sth triggered.. > > > could you pastebin the RS log? > > > > > > On Wed, Sep 10, 2014 at 6:59 PM, Brian Jeltema < > > brian.jelt...@digitalenvoy.net> wrote: > > > > out of curiosity, did you see below messages in RS log? > > > LOG.warn("Snapshot called again without clearing previous. " + > > "Doing nothing. Another ongoing flush or did we fail last > > attempt?”); > > > > Nope > > > > > thanks. > > > On Tue, Sep 9, 2014 at 2:15 AM, Brian Jeltema < > > brian.jelt...@digitalenvoy.net> wrote: > > > I’ve resolved these problems by restarting the region server that owned > > the region in question. > > I don’t know what the underlying issue was, but at this point it’s not > > worth pursuing. > > > Thanks for responding. > > > Brian > > > On Sep 8, 2014, at 11:06 AM, Brian Jeltema < > > brian.jelt...@digitalenvoy.net> > > wrote: > > > I realized today that the region server logs for the region being > > updated (startKey=\x00DDD@) contains the following: > > > 2014-09-08 06:25:50,223 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 11302 > > 2014-09-08 06:26:00,222 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 21682 > > 2014-09-08 06:26:10,223 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 5724 > > 2014-09-08 06:26:20,223 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 11962 > > 2014-09-08 06:26:30,223 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 7693 > > 2014-09-08 06:26:40,224 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 5578 > > 2014-09-08 06:26:50,223 INFO [regionserver60020.periodicFlusher] > > regionserver.HRegionServer: regionserver60020.periodicFlusher requesting > > flush for region Host,\x00DDD@ > > ,1400624237999.5bb6bd41597ddd8dd7ca03e78f3a3e65. > > after a delay of 12420 > > > a log entry being generated every 10 seconds starting about 4 days ago. > > I presume these problems are related. > > > On Sep 8, 2014, at 7:10 AM, Brian Jeltema < > > brian.jelt...@digitalenvoy.net> wrote: > > > > When number of attempts is greater than the value of > > hbase.client.start.log.errors.counter (default 9), AsyncProcess would > > produce logs cited below. > > The interval following 'retrying after ' is the backoff time. > > > Which release of HBase are you using ? > > > > HBase Version 0.98.0.2.1.1.0-385-hadoop2 > > > The MR job is reading from an HBase snapshot, if that’s relevant. > > > Cheers > > > > On Sun, Sep 7, 2014 at 8:50 AM, Brian Jeltema < > > brian.jelt...@digitalenvoy.net> wrote: > > > I have a map/reduce job that is consistently failing with timeouts. > > The > > failing mapper log files contain a series > > of records similar to those below. When I look at the hbase and hdfs > > logs > > (on foo.net in this case) I don’t see > > anything obvious at these timestamps. The mapper task times out > > at/near > > attempt=25/35. Can anyone shed light > > on what these log entries mean? > > > Thanks - Brian > > > > 2014-09-07 09:36:51,421 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=10/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 10029 ms, replay 1062 ops > > 2014-09-07 09:37:01,642 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=11/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 10023 ms, replay 1062 ops > > 2014-09-07 09:37:12,064 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=12/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 20182 ms, replay 1062 ops > > 2014-09-07 09:37:32,708 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=13/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 20140 ms, replay 1062 ops > > 2014-09-07 09:37:52,940 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=14/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 20041 ms, replay 1062 ops > > 2014-09-07 09:38:13,324 INFO [htable-pool1-t1] > > org.apache.hadoop.hbase.client.AsyncProcess: #3, table=Host, > > primary, > > attempt=15/35 failed 1062 ops, last exception: null on foo.net > > ,60020,1406043467187, > > tracking started null, retrying after 20041 ms, replay 1062 ops
