[
https://issues.apache.org/jira/browse/KYLIN-2749?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16098088#comment-16098088
]
Alexander Sterligov commented on KYLIN-2749:
--------------------------------------------
It is reproduced again. It is caused by kylin OOM (system, not OutOfMemoryError)
Merge process is in progress, old segment tables are already removed, but cube
json contains references for them.
Merged segment is in NEW state and "to hfile" job is running.
It looks like it started job, which successfully worked almost up to the end
and then started a new one after OOM.
Does kylin run its jobs in separate process?
> Merge may remove old segments without saving merged segment
> -----------------------------------------------------------
>
> Key: KYLIN-2749
> URL: https://issues.apache.org/jira/browse/KYLIN-2749
> Project: Kylin
> Issue Type: Bug
> Affects Versions: v2.0.0
> Reporter: Alexander Sterligov
>
> Merge started to work on last 7 segments.
> During the process hbase had a failure because of spot-instances shutdown in
> Amazon. Data was not lost, because it is at S3.
> I stopped kylin and did hbase hbck --repair. During the report of repair I
> didn't see any information about lost data, just redistribution of regions.
> Then after kylin was started I cannot query data from the last 7 segments:
> {quote}
> Caused by: java.lang.RuntimeException:
> org.apache.hadoop.hbase.TableNotFoundException: Table 'KYLIN_7MMHCHKVVB' was
> not found, got: KYLIN_7H3WSPX1UJ.
> at com.google.common.base.Throwables.propagate(Throwables.java:160)
> at
> org.apache.kylin.storage.hbase.cube.v2.ExpectedSizeIterator.next(ExpectedSizeIterator.java:67)
> at
> org.apache.kylin.storage.hbase.cube.v2.ExpectedSizeIterator.next(ExpectedSizeIterator.java:31)
> at
> com.google.common.collect.TransformedIterator.next(TransformedIterator.java:48)
> at com.google.common.collect.Iterators$6.hasNext(Iterators.java:583)
> at
> org.apache.kylin.storage.gtrecord.SegmentCubeTupleIterator$2.hasNext(SegmentCubeTupleIterator.java:116)
> at
> org.apache.kylin.storage.gtrecord.SegmentCubeTupleIterator.hasNext(SegmentCubeTupleIterator.java:149)
> at com.google.common.collect.Iterators$6.hasNext(Iterators.java:582)
> at
> org.apache.kylin.storage.gtrecord.SequentialCubeTupleIterator.hasNext(SequentialCubeTupleIterator.java:129)
> at
> org.apache.kylin.query.enumerator.OLAPEnumerator.moveNext(OLAPEnumerator.java:67)
> at Baz$1$1.moveNext(Unknown Source)
> at
> org.apache.calcite.linq4j.EnumerableDefaults.groupBy_(EnumerableDefaults.java:826)
> at
> org.apache.calcite.linq4j.EnumerableDefaults.groupBy(EnumerableDefaults.java:761)
> at
> org.apache.calcite.linq4j.DefaultEnumerable.groupBy(DefaultEnumerable.java:302)
> at Baz.bind(Unknown Source)
> at
> org.apache.calcite.jdbc.CalcitePrepare$CalciteSignature.enumerable(CalcitePrepare.java:331)
> at
> org.apache.calcite.jdbc.CalciteConnectionImpl.enumerable(CalciteConnectionImpl.java:294)
> at
> org.apache.calcite.jdbc.CalciteMetaImpl._createIterable(CalciteMetaImpl.java:553)
> at
> org.apache.calcite.jdbc.CalciteMetaImpl.createIterable(CalciteMetaImpl.java:544)
> at
> org.apache.calcite.avatica.AvaticaResultSet.execute(AvaticaResultSet.java:193)
> at
> org.apache.calcite.jdbc.CalciteResultSet.execute(CalciteResultSet.java:67)
> at
> org.apache.calcite.jdbc.CalciteResultSet.execute(CalciteResultSet.java:44)
> at
> org.apache.calcite.avatica.AvaticaConnection$1.execute(AvaticaConnection.java:607)
> at
> org.apache.calcite.jdbc.CalciteMetaImpl.prepareAndExecute(CalciteMetaImpl.java:600)
> at
> org.apache.calcite.avatica.AvaticaConnection.prepareAndExecuteInternal(AvaticaConnection.java:615)
> at
> org.apache.calcite.avatica.AvaticaStatement.executeInternal(AvaticaStatement.java:148)
> ... 77 more
> Caused by: org.apache.hadoop.hbase.TableNotFoundException: Table
> 'KYLIN_7MMHCHKVVB' was not found, got: KYLIN_7H3WSPX1UJ.
> at
> org.apache.hadoop.hbase.client.ConnectionManager$HConnectionImplementation.locateRegionInMeta(ConnectionManager.java:1310)
> at
> org.apache.hadoop.hbase.client.ConnectionManager$HConnectionImplementation.locateRegion(ConnectionManager.java:1189)
> at
> org.apache.hadoop.hbase.client.ConnectionManager$HConnectionImplementation.locateRegion(ConnectionManager.java:1173)
> at
> org.apache.hadoop.hbase.client.ConnectionManager$HConnectionImplementation.locateRegion(ConnectionManager.java:1130)
> at
> org.apache.hadoop.hbase.client.ConnectionManager$HConnectionImplementation.getRegionLocation(ConnectionManager.java:965)
> at
> org.apache.hadoop.hbase.client.HRegionLocator.getRegionLocation(HRegionLocator.java:83)
> at
> org.apache.hadoop.hbase.client.HTable.getRegionLocation(HTable.java:505)
> at
> org.apache.hadoop.hbase.client.HTable.getKeysAndRegionsInRange(HTable.java:721)
> at
> org.apache.hadoop.hbase.client.HTable.getKeysAndRegionsInRange(HTable.java:691)
> at
> org.apache.hadoop.hbase.client.HTable.getStartKeysInRange(HTable.java:1796)
> at
> org.apache.hadoop.hbase.client.HTable.coprocessorService(HTable.java:1751)
> at
> org.apache.kylin.storage.hbase.cube.v2.CubeHBaseEndpointRPC$1.run(CubeHBaseEndpointRPC.java:182)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
> ... 1 more
> {quote}
> hbase really doesn't contain tables for the last 7 segments and I didn't call
> any cleanup jobs.
> It looks like Merge removed old segment tables before merged segment was
> saved.
> I'm going to continue to investigate this problem and will post more details
> next week.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
