[
https://issues.apache.org/jira/browse/CASSANDRA-12730?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15645739#comment-15645739
]
Kurt Greaves commented on CASSANDRA-12730:
------------------------------------------
For the record, it seems to me like this is a potentially harmless
optimisation, but doesn't really fix the core issue.
I've seen similar issues on a lot of clusters on a lot of different versions
(pretty much every version since 2.1.11). This issue occurs quite frequently
and not necessarily on clusters under heap pressure. Usually when we see this
issue the logs report masses of streaming sessions and flushes. The proposed
solution seems to be simple and harmless enough even though it may not fix the
real issue and I think would help in cases where there are a small amount of
constant writes to a table such that the memtable is never/rarely clean when
the flush is triggered, resulting in lots of small SSTables.
I think the real issue is a problem elsewhere that triggers lots of stream
sessions during the repair even though the data is consistent. We've seen this
sort of thing in cases where we're fairly sure there have been no down nodes
and no dropped mutations, and even running repairs continuously results in many
streams.
> Thousands of empty SSTables created during repair - TMOF death
> --------------------------------------------------------------
>
> Key: CASSANDRA-12730
> URL: https://issues.apache.org/jira/browse/CASSANDRA-12730
> Project: Cassandra
> Issue Type: Bug
> Components: Local Write-Read Paths
> Reporter: Benjamin Roth
> Priority: Critical
>
> Last night I ran a repair on a keyspace with 7 tables and 4 MVs each
> containing a few hundret million records. After a few hours a node died
> because of "too many open files".
> Normally one would just raise the limit, but: We already set this to 100k.
> The problem was that the repair created roughly over 100k SSTables for a
> certain MV. The strange thing is that these SSTables had almost no data (like
> 53bytes, 90bytes, ...). Some of them (<5%) had a few 100 KB, very few (<1%
> had normal sizes like >= few MB). I could understand, that SSTables queue up
> as they are flushed and not compacted in time but then they should have at
> least a few MB (depending on config and avail mem), right?
> Of course then the node runs out of FDs and I guess it is not a good idea to
> raise the limit even higher as I expect that this would just create even more
> empty SSTables before dying at last.
> Only 1 CF (MV) was affected. All other CFs (also MVs) behave sanely. Empty
> SSTables have been created equally over time. 100-150 every minute. Among the
> empty SSTables there are also Tables that look normal like having few MBs.
> I didn't see any errors or exceptions in the logs until TMOF occured. Just
> tons of streams due to the repair (which I actually run over cs-reaper as
> subrange, full repairs).
> After having restarted that node (and no more repair running), the number of
> SSTables went down again as they are compacted away slowly.
> According to [~zznate] this issue may relate to CASSANDRA-10342 +
> CASSANDRA-8641
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
