[
https://issues.apache.org/jira/browse/AURORA-1481?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14735473#comment-14735473
]
Paul Cavallaro commented on AURORA-1481:
----------------------------------------
Qualitatively it also makes the thermos observer web ui, fairly unresponsive,
i.e., waiting multiple seconds for viewing a task.
> Thermos Observer Takes up to 150% CPU on default settings
> ---------------------------------------------------------
>
> Key: AURORA-1481
> URL: https://issues.apache.org/jira/browse/AURORA-1481
> Project: Aurora
> Issue Type: Bug
> Affects Versions: 0.9.0
> Environment: CentOS 6.6 w/ Mesos 0.22.0
> Reporter: Paul Cavallaro
> Priority: Minor
>
> The Thermos Observer when default configured with 5 second polling interval
> takes up to 150% cpu on an aws c3.4xlarge. This is with
> ~10k executors
> ~1 run per executor
> Haven't done any profiling, but the overall design seems to be inefficient
> given it has to list read the entire directory tree every time and do a fair
> amount of regex parsing.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
