[
https://issues.apache.org/jira/browse/YARN-138?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siddharth Seth updated YARN-138:
--------------------------------
Attachment: YARN138.txt
Re-based Harsh's patch, changed defaults min-allocation to 1024, and default
max-allocation to 8192. NM resource.memory-mb default is already at 8192.
With this (and with the old config) - a default install allows 1 AM + a total
of 6 map/reduce tasks. The M/R slots will not go via reservations in the
scheduler though.
bq. Let's just bump up default to 1024, 512 is still too low for boxes with 16G
of RAM and 8 cores.
Agree with bumping it up to 1024 - at least to reduce the default number of
containers per box. Maybe we should consider reducing the default NM
available-mb as well for default installs. 8 containers may be too high for
someone just trying YARN.
> Improve default config values for YARN
> --------------------------------------
>
> Key: YARN-138
> URL: https://issues.apache.org/jira/browse/YARN-138
> Project: Hadoop YARN
> Issue Type: Bug
> Components: resourcemanager, scheduler
> Affects Versions: 2.0.0-alpha
> Reporter: Arun C Murthy
> Assignee: Harsh J
> Labels: performance
> Attachments: MAPREDUCE-4316.patch, YARN138.txt
>
>
> Currently some of our configs are way off e.g. min-alloc is 128M while
> max-alloc is 10240.
> This leads to poor out-of-box performance as noticed by some users:
> http://s.apache.org/avd
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
