[
https://issues.apache.org/jira/browse/HADOOP-1984?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12543057
]
Amar Kamat commented on HADOOP-1984:
------------------------------------
Consider the following strategy :
{code}
backoff (t) = backoff-init * backoff-base^{t-1}
t = number of retries
backoff-init = 4 sec
backoff-base = 2
max-t = 5
{code}
So now the backoff intervals are {{4, 8, 16, 32, 64, 128}}, {{sum = 252 = 4.2
min}} after which the notification is sent to the jobtracker from the reduce.
So 6 retries are made before sending the notification.
> some reducer stuck at copy phase and progress extremely slowly
> --------------------------------------------------------------
>
> Key: HADOOP-1984
> URL: https://issues.apache.org/jira/browse/HADOOP-1984
> Project: Hadoop
> Issue Type: Bug
> Components: mapred
> Affects Versions: 0.16.0
> Reporter: Runping Qi
> Assignee: Amar Kamat
> Priority: Critical
> Fix For: 0.16.0
>
> Attachments: HADOOP-1984.patch
>
>
> In many cases, some reducers got stuck at copy phase, progressing extremely
> slowly.
> The entire cluster seems doing nothing. This causes a very bad long tails of
> otherwise well tuned map/red jobs.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
