hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Nicolas Liochon (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-9775) Client write path perf issues
Date Wed, 16 Oct 2013 13:49:44 GMT

    [ https://issues.apache.org/jira/browse/HBASE-9775?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13796798#comment-13796798

Nicolas Liochon commented on HBASE-9775:

On the perfs, beyond possible bugs, I see:
- with a max.total.tasks of 100 and max.perserver.tasks of 5, the client might not use all
the server. May be a default of 2 for max.perserver.tasks would be better
- with 32 clients and 77 servers, the client can have 32*77 threads (~2400) but in 0.96 we
have a default limit of 256. May be we should increase it to something like 1024? We're not
async on the client side, so...

For the failure, I don"t know. Something I'm not totally sure about is HBASE-9467: we're now
rejecting the client when we're busy, instead of keeping it in the queue. A possible side
effect when the system is heavily loaded with a lot of client is that the rejected client
is always unlucky, and its next attempts are rejected as well: wedon't have any priority between
the new callers and the old callers that were rejected.

> Client write path perf issues
> -----------------------------
>                 Key: HBASE-9775
>                 URL: https://issues.apache.org/jira/browse/HBASE-9775
>             Project: HBase
>          Issue Type: Bug
>          Components: Client
>    Affects Versions: 0.96.0
>            Reporter: Elliott Clark
>            Priority: Critical
>         Attachments: Charts Search   Cloudera Manager.png, short_ycsb.png
> Testing on larger clusters has not had the desired throughput increases.

This message was sent by Atlassian JIRA

View raw message