[
https://issues.apache.org/jira/browse/HADOOP-12421?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15224972#comment-15224972
]
Jing Zhao edited comment on HADOOP-12421 at 4/4/16 8:21 PM:
------------------------------------------------------------
Thanks for working on this, [~eclark]. One quick question is currently the
client's retry policy in HA setup, {{FailoverOnNetworkExceptionRetry}}, already
includes randomness in {{calculateExponentialTime}}. Do you know why this
cannot avoid the spike and why an extra jitter can help here?
was (Author: jingzhao):
Thanks for working on this, [~eclark]. One quick question is currently the
client's retry policy, {{FailoverOnNetworkExceptionRetry}}, already includes
randomness in {{calculateExponentialTime}}. Do you know why this cannot avoid
the spike and why an extra jitter can help here?
> Add jitter to RetryInvocationHandler
> ------------------------------------
>
> Key: HADOOP-12421
> URL: https://issues.apache.org/jira/browse/HADOOP-12421
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 2.8.0
> Reporter: Elliott Clark
> Assignee: Elliott Clark
> Attachments: HADOOP-12421-v1.patch, HADOOP-12421-v2.patch,
> HADOOP-12421-v3.patch, HADOOP-12421-v4.patch, HADOOP-12421-v5.patch
>
>
> Calls to NN can become synchronized across a cluster during NN failover. This
> leads to a spike in requests until things recover. Making an already tricky
> time worse.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
