[
https://issues.apache.org/jira/browse/HBASE-26067?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17468264#comment-17468264
]
Josh Elser commented on HBASE-26067:
------------------------------------
Ran a simple {{hbase pe randomWrite}} and I'm noticing one thing that we also
saw when backporting this to branch-2.4 (internally)
{noformat}
2022-01-03 17:40:47,184 INFO [MemStoreFlusher.1]
regionserver.DefaultStoreFlusher(81): Flushed memstore data size=73.26 MB at
sequenceid=4537 (bloomFilter=true),
to=hdfs://mizar.cloudera:8020/hbase-2.6/data/default/TestTable/951d5e954f95ab58f224fe80a77bea56/info0/7732094
d69a94641a56965c4fb0d1947 {noformat}
and
{noformat}
2022-01-03 17:41:18,319 ERROR [regionserver/localhost:16020-shortCompactions-0]
regionserver.CompactSplit(670): Compaction failed
region=TestTable,00000000000000000004054490,1641249249856.2dc7251c6eceb660b9c7bb0b587db913.,
storeName=2dc7251c6eceb660b9c7bb0b587db913/info0, priority=6,
startTime=1641249666161
java.io.IOException: Root-level entries already added in single-level mode
at
org.apache.hadoop.hbase.io.hfile.HFileBlockIndex$BlockIndexWriter.writeSingleLevelIndex(HFileBlockIndex.java:1136)
at
org.apache.hadoop.hbase.io.hfile.CompoundBloomFilterWriter$MetaWriter.write(CompoundBloomFilterWriter.java:279)
at
org.apache.hadoop.hbase.io.hfile.HFileWriterImpl$1.writeToBlock(HFileWriterImpl.java:713)
at
org.apache.hadoop.hbase.io.hfile.HFileBlock$Writer.writeBlock(HFileBlock.java:1205)
at
org.apache.hadoop.hbase.io.hfile.HFileWriterImpl.close(HFileWriterImpl.java:660)
at
org.apache.hadoop.hbase.regionserver.StoreFileWriter.close(StoreFileWriter.java:377)
at
org.apache.hadoop.hbase.regionserver.compactions.DefaultCompactor.commitWriter(DefaultCompactor.java:70)
at
org.apache.hadoop.hbase.regionserver.compactions.Compactor.compact(Compactor.java:386)
at
org.apache.hadoop.hbase.regionserver.compactions.DefaultCompactor.compact(DefaultCompactor.java:62)
at
org.apache.hadoop.hbase.regionserver.DefaultStoreEngine$DefaultCompactionContext.compact(DefaultStoreEngine.java:125)
at org.apache.hadoop.hbase.regionserver.HStore.compact(HStore.java:1141)
at org.apache.hadoop.hbase.regionserver.HRegion.compact(HRegion.java:2388)
at
org.apache.hadoop.hbase.regionserver.CompactSplit$CompactionRunner.doCompaction(CompactSplit.java:654)
at
org.apache.hadoop.hbase.regionserver.CompactSplit$CompactionRunner.run(CompactSplit.java:697)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748) {noformat}
We had figured out that this started happening when we backported HBASE-26271.
[~bszabolcs] actually realized that we had transposed the arguments in
{{sinkFactory.createWriter}} (substituting the dropCache for
request.isMajor()). I'm looking to see if I can spot something obvious, but
maybe the same thing happened here.
> Change the way on how we track store file list
> ----------------------------------------------
>
> Key: HBASE-26067
> URL: https://issues.apache.org/jira/browse/HBASE-26067
> Project: HBase
> Issue Type: Umbrella
> Components: HFile
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Priority: Major
> Fix For: 3.0.0-alpha-3
>
>
> Open a separated jira to track the work since it can not be fully included in
> HBASE-24749.
> I think this could be a landed prior to HBASE-24749, as if this works, we
> could have different implementations for tracking store file list.
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
