cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Robert Stupp (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-13754) FastThreadLocal leaks memory
Date Thu, 31 Aug 2017 16:25:00 GMT


Robert Stupp commented on CASSANDRA-13754:

Well, yea. Looking at the heap dump, that [~markusdlugi] provided, is looks like the node
is "just" overloaded with too many and maybe too big writes in combination with a small heap.
There are lots of {{BTree$Builder}} instances with live references in their {{Object[] values}}
array to {{HeapByteBuffer}} instances, each holding a 1MB {{byte[]}}.
{{BTree$Builder}} instances reset the {{Object[] values}} when finished - i.e. those builders
are actively doing something (writes are happening at that time).
TL;DR I don't think this is actually related to the issue that [~urandom] describes.

[~urandom], can you explain what actually what these {{ThreadLocal}} instances referenced?

> FastThreadLocal leaks memory
> ----------------------------
>                 Key: CASSANDRA-13754
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>         Environment: Cassandra 3.11.0, Netty 4.0.44.Final, OpenJDK 8u141-b15
>            Reporter: Eric Evans
>            Assignee: Robert Stupp
>             Fix For: 3.11.1
> After a chronic bout of {{OutOfMemoryError}} in our development environment, a heap analysis
is showing that more than 10G of our 12G heaps are consumed by the {{threadLocals}} members
(instances of {{java.lang.ThreadLocalMap}}) of various {{io.netty.util.concurrent.FastThreadLocalThread}}
instances.  Reverting [cecbe17|;a=commit;h=cecbe17e3eafc052acc13950494f7dddf026aa54]
fixes the issue.

This message was sent by Atlassian JIRA

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message