[
https://issues.apache.org/jira/browse/CASSANDRA-5661?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13707988#comment-13707988
]
Pavel Yaskevich edited comment on CASSANDRA-5661 at 7/14/13 8:53 AM:
---------------------------------------------------------------------
Finalizers are never a good idea, because it requires double pass though GC and
secondary actual finalization processing is single threaded with global queue,
all of which creates additional CPU overhead as well as memory pressure. Also
as as far as I remember CMS/G1 both process all types of references on FullGC
phase even if refs to underlying objects where lost in young gen they are still
promoted to old gen.
was (Author: xedin):
Finalizers are never a good idea, because it requires double pass though GC
and secondary actual finalization processing is single threaded with global
queue, all of which creates additional CPU overhead as well as memory pressure.
> Discard pooled readers for cold data
> ------------------------------------
>
> Key: CASSANDRA-5661
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5661
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 1.2.1
> Reporter: Jonathan Ellis
> Assignee: Pavel Yaskevich
> Fix For: 2.0
>
> Attachments: CASSANDRA-5661-multiway-per-sstable.patch,
> CASSANDRA-5661.patch, CASSANDRA-5661-v2-global-multiway-per-sstable.patch,
> DominatorTree.png, Histogram.png
>
>
> Reader pooling was introduced in CASSANDRA-4942 but pooled
> RandomAccessReaders are never cleaned up until the SSTableReader is closed.
> So memory use is "the worst case simultaneous RAR we had open for this file,
> forever."
> We should introduce a global limit on how much memory to use for RAR, and
> evict old ones.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
