hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Chang Li (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (HADOOP-12645) Add option to allow multiple threads making calls on same RPC client not retry serially
Date Tue, 15 Dec 2015 22:35:46 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-12645?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Chang Li updated HADOOP-12645:
------------------------------
    Attachment: HADOOP-12645.1.patch

upload a patch to illustrate this idea. will try to work out a unit test.

> Add option to allow multiple threads making calls on same RPC client not retry serially
> ---------------------------------------------------------------------------------------
>
>                 Key: HADOOP-12645
>                 URL: https://issues.apache.org/jira/browse/HADOOP-12645
>             Project: Hadoop Common
>          Issue Type: Improvement
>            Reporter: Chang Li
>            Assignee: Chang Li
>         Attachments: HADOOP-12645.1.patch
>
>
> Current multiple threads making calls on same RPC client retry serially. For example,
when a first call starts for retrying for 10 minutes, second and third and fourth calls called
in during this time will get queued up. After 1st call fail, those second, third and fourth
call will still unwisely go ahead serially to do the 10 minutes retry and fail. Propose to
add a optional setting, default to false, but when enabled will allow those queued up following
calls to fail without wasting time to retry.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message