[
https://issues.apache.org/jira/browse/HBASE-20539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16467667#comment-16467667
]
stack commented on HBASE-20539:
-------------------------------
Having a bit of trouble with this one. Locally wth unit test all looks good but
out on my cluster I am not picking up the default NONE in hbase-default.xml...
and so see stuff like this:
25453 2018-05-08 09:34:11,871 INFO
[StoreOpener-88eb562860c251ec625f301da56e4123-1] regionserver.HStore:
inMemoryCompactionPolicy=BASIC, ms=class
org.apache.hadoop.hbase.regionserver.CompactingMemStore
Or, running w/ a bad patch that throws NPEs... I see we are referring to
compactingmemstore...
{code}
1198945 java.lang.NullPointerException
1198946 at
org.apache.hadoop.hbase.util.CollectionBackedScanner.init(CollectionBackedScanner.java:77)
1198947 at
org.apache.hadoop.hbase.util.CollectionBackedScanner.<init>(CollectionBackedScanner.java:50)
1198948 at
org.apache.hadoop.hbase.regionserver.Segment.getScannersForSnapshot(Segment.java:141)
1198949 at
org.apache.hadoop.hbase.regionserver.MemStoreSnapshot.<init>(MemStoreSnapshot.java:44)
1198950 at
org.apache.hadoop.hbase.regionserver.CompactingMemStore.snapshot(CompactingMemStore.java:216)
1198951 at
org.apache.hadoop.hbase.regionserver.HStore$StoreFlusherImpl.prepare(HStore.java:2203)
1198952 at
org.apache.hadoop.hbase.regionserver.HRegion.lambda$internalPrepareFlushCache$2(HRegion.java:2493)
1198953 at java.util.TreeMap.forEach(TreeMap.java:1005)
1198954 at
org.apache.hadoop.hbase.regionserver.HRegion.internalPrepareFlushCache(HRegion.java:2492)
1198955 at
org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRegion.java:2364)
1198956 at
org.apache.hadoop.hbase.regionserver.HRegion.internalFlushcache(HRegion.java:2338)
1198957 at
org.apache.hadoop.hbase.regionserver.HRegion.flushcache(HRegion.java:2229)
1198958 at
org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushRegion(MemStoreFlusher.java:607)
1198959 at
org.apache.hadoop.hbase.regionserver.MemStoreFlusher.flushOneForGlobalPressure(MemStoreFlusher.java:290)
1198960 at
org.apache.hadoop.hbase.regionserver.MemStoreFlusher.access$700(MemStoreFlusher.java:68)
1198961 at
org.apache.hadoop.hbase.regionserver.MemStoreFlusher$FlushHandler.run(MemStoreFlusher.java:346)
1198962 at java.lang.Thread.run(Thread.java:748)
{code}
It may be just an issue on my cluster... but I'm running tip of branch-2.0.
> Disable IMC; part 2
> -------------------
>
> Key: HBASE-20539
> URL: https://issues.apache.org/jira/browse/HBASE-20539
> Project: HBase
> Issue Type: Sub-task
> Components: in-memory-compaction
> Reporter: stack
> Assignee: stack
> Priority: Major
> Fix For: 2.0.1
>
>
> Just noticed that post-flush, we are picking up a CompactingMemStore though
> IMC is supposed to be off. We are picking up an inMemoryCompaction of BASIC
> (noticed a day or so ago by [~chia7712]) and so, we fall into the default in
> HStore#getMemstore which is set to be a CompactingMemStore....
> No real perf difference going by the compares I've been running but let me
> clean this up.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
