[
https://issues.apache.org/jira/browse/MESOS-3078?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14727841#comment-14727841
]
Benjamin Mahler commented on MESOS-3078:
----------------------------------------
MESOS-3157 is not related to this, only performing batch allocations reduces
throughput per the discussion in the email thread linked in the description,
and is a workaround for known performance issues.
The idea here is to implement (1) per Klaus' suggestion, however we need to be
careful about responsiveness given the known performance issues in the
allocation loop (see MESOS-3283).
> 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
>
> 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)
