[
https://issues.apache.org/jira/browse/HBASE-17738?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16094489#comment-16094489
]
Anoop Sam John commented on HBASE-17738:
----------------------------------------
So as per this, the 15 threads will create 1 each BB and the last will create
10?
The test case issue came because we were creating so less BBs and available
processors are more. May be we should be having a min BBs that each threads
should create? May be 10. We should create only those many threads then. I
feel that should be the correct approach than this Math.floor() way which will
create large diff btw #BBs created by last threads vs all others. Say when
99 BBs to create and 10 threads, all 9 threads will create 9 BBs each where as
the last will create 18
> BucketCache startup is slow
> ---------------------------
>
> Key: HBASE-17738
> URL: https://issues.apache.org/jira/browse/HBASE-17738
> Project: HBase
> Issue Type: Sub-task
> Components: BucketCache
> Affects Versions: 2.0.0
> Reporter: stack
> Assignee: ramkrishna.s.vasudevan
> Fix For: 2.0.0, 2.0.0-alpha-2
>
> Attachments: HBASE-17738_10.patch, HBASE-17738_2.patch,
> HBASE-17738_2.patch, HBASE-17738_3.patch, HBASE-17738_4.patch,
> HBASE-17738_5_withoutUnsafe.patch, HBASE-17738_6_withoutUnsafe.patch,
> HBASE-17738_8.patch, HBASE-17738_9.patch, HBASE-17738_log.patch,
> HBASE-17738.patch
>
>
> If you set bucketcache size at 64G say and then start hbase, it takes a long
> time. Can we do the allocations in parallel and not inline with the server
> startup?
> Related, prefetching on a bucketcache is slow. Speed it up.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
