[
https://issues.apache.org/jira/browse/CASSANDRA-6181?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13826340#comment-13826340
]
Sylvain Lebresne commented on CASSANDRA-6181:
---------------------------------------------
This is related to range tombstones. We mainly use range tombstone in two cases:
# you do a 'DELETE FROM X WHERE Y' (so the whole CQL3 row) *and* Y contains at
least a clustering column (if it contains only the partition key for instance,
we don't use a range tombstone).
# you set a collection (but not when you append/add/remove to it).
And this may be triggered if you reach some amount of such range tombstones
within one partition. But 1) this is not guaranteed at all, as this depends on
the ranges covered by those tombstones and the insertion order of those
tombstones (and the exact conditions can't be easily characterize, sorry) and
2) the exact number of tombstones after which this may be triggered depends on
the stack size (see Fabien's comment above).
In any case, the patch eliminate this entirely.
> Replaying a commit led to java.lang.StackOverflowError and node crash
> ---------------------------------------------------------------------
>
> Key: CASSANDRA-6181
> URL: https://issues.apache.org/jira/browse/CASSANDRA-6181
> Project: Cassandra
> Issue Type: Bug
> Environment: 1.2.8 & 1.2.10 - ubuntu 12.04
> Reporter: Jeffrey Damick
> Assignee: Sylvain Lebresne
> Priority: Critical
> Fix For: 1.2.12, 2.0.2
>
> Attachments: 6181.txt
>
>
> 2 of our nodes died after attempting to replay a commit. I can attach the
> commit log file if that helps.
> It was occurring on 1.2.8, after several failed attempts to start, we
> attempted startup with 1.2.10. This also yielded the same issue (below).
> The only resolution was to physically move the commit log file out of the way
> and then the nodes were able to start...
> The replication factor was 3 so I'm hoping there was no data loss...
> {code}
> INFO [main] 2013-10-11 14:50:35,891 CommitLogReplayer.java (line 119)
> Replaying /ebs/cassandra/commitlog/CommitLog-2-1377542389560.log
> ERROR [MutationStage:18] 2013-10-11 14:50:37,387 CassandraDaemon.java (line
> 191) Exception in thread Thread[MutationStage:18,5,main]
> java.lang.StackOverflowError
> at
> org.apache.cassandra.db.marshal.TimeUUIDType.compareTimestampBytes(TimeUUIDType.java:68)
> at
> org.apache.cassandra.db.marshal.TimeUUIDType.compare(TimeUUIDType.java:57)
> at
> org.apache.cassandra.db.marshal.TimeUUIDType.compare(TimeUUIDType.java:29)
> at
> org.apache.cassandra.db.marshal.AbstractType.compareCollectionMembers(AbstractType.java:229)
> at
> org.apache.cassandra.db.marshal.AbstractCompositeType.compare(AbstractCompositeType.java:81)
> at
> org.apache.cassandra.db.marshal.AbstractCompositeType.compare(AbstractCompositeType.java:31)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertAfter(RangeTombstoneList.java:439)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertFrom(RangeTombstoneList.java:405)
> at
> org.apache.cassandra.db.RangeTombstoneList.weakInsertFrom(RangeTombstoneList.java:472)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertAfter(RangeTombstoneList.java:456)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertFrom(RangeTombstoneList.java:405)
> at
> org.apache.cassandra.db.RangeTombstoneList.weakInsertFrom(RangeTombstoneList.java:472)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertAfter(RangeTombstoneList.java:456)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertFrom(RangeTombstoneList.java:405)
> at
> org.apache.cassandra.db.RangeTombstoneList.weakInsertFrom(RangeTombstoneList.java:472)
> .... etc.... over and over until ....
> at
> org.apache.cassandra.db.RangeTombstoneList.weakInsertFrom(RangeTombstoneList.java:472)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertAfter(RangeTombstoneList.java:456)
> at
> org.apache.cassandra.db.RangeTombstoneList.insertFrom(RangeTombstoneList.java:405)
> at
> org.apache.cassandra.db.RangeTombstoneList.add(RangeTombstoneList.java:144)
> at
> org.apache.cassandra.db.RangeTombstoneList.addAll(RangeTombstoneList.java:186)
> at org.apache.cassandra.db.DeletionInfo.add(DeletionInfo.java:180)
> at
> org.apache.cassandra.db.AtomicSortedColumns.addAllWithSizeDelta(AtomicSortedColumns.java:197)
> at
> org.apache.cassandra.db.AbstractColumnContainer.addAllWithSizeDelta(AbstractColumnContainer.java:99)
> at org.apache.cassandra.db.Memtable.resolve(Memtable.java:207)
> at org.apache.cassandra.db.Memtable.put(Memtable.java:170)
> at
> org.apache.cassandra.db.ColumnFamilyStore.apply(ColumnFamilyStore.java:745)
> at org.apache.cassandra.db.Table.apply(Table.java:388)
> at org.apache.cassandra.db.Table.apply(Table.java:353)
> at
> org.apache.cassandra.db.commitlog.CommitLogReplayer$1.runMayThrow(CommitLogReplayer.java:258)
> at
> org.apache.cassandra.utils.WrappedRunnable.run(WrappedRunnable.java:28)
> at
> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:471)
> at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:334)
> at java.util.concurrent.FutureTask.run(FutureTask.java:166)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)
> at java.lang.Thread.run(Thread.java:724)
> {code}
--
This message was sent by Atlassian JIRA
(v6.1#6144)
