[
https://issues.apache.org/jira/browse/MESOS-3078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15116856#comment-15116856
]
Klaus Ma commented on MESOS-3078:
---------------------------------
One more idea for this JIRA: I think we can let framework keep using the offer
until master/allocator decide to reclaim/re-shuffle it, it'll increase the
throughput of cluster and reduce the idle time of resources in allocator.
> Recovered resources are not re-allocated until the next allocation delay.
> -------------------------------------------------------------------------
>
> Key: MESOS-3078
> URL: https://issues.apache.org/jira/browse/MESOS-3078
> Project: Mesos
> Issue Type: Improvement
> Components: allocation
> Reporter: Benjamin Mahler
> Assignee: Klaus Ma
>
> Currently, when resources are recovered, we do not perform an allocation for
> that slave. Rather, we wait until the next allocation interval.
> For small task, high throughput frameworks, this can have a significant
> impact on overall throughput, see the following thread:
> http://markmail.org/thread/y6mzfwzlurv6nik3
> We should consider immediately performing a re-allocation for the slave upon
> resource recovery.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)