[
https://issues.apache.org/jira/browse/HBASE-9775?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13799222#comment-13799222
]
Jeffrey Zhong commented on HBASE-9775:
--------------------------------------
[~nkeywal] You're right. It's not real recursive. The task counter double bump
up time is very short depends on how long the runnable are resubmitted.
While we still have issues, when there are failures which will stay in the
retry queue and still use task counters even during the sleep interval for
retry. In Elliot testis, there exists retry failures so I think that blocked
some new edits to be sent out. In the posted patch, we only bump up task
counter during the time when tasks are being sent out and waiting for response.
> 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 - ITBLL.png, Charts
> Search Cloudera Manager.png, hbase-9775.patch, job_run.log, short_ycsb.png,
> ycsb_insert_94_vs_96.png
>
>
> Testing on larger clusters has not had the desired throughput increases.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
