[ https://issues.apache.org/jira/browse/HDDS-15093 ]
Wei-Chiu Chuang deleted comment on HDDS-15093:
----------------------------------------
was (Author: jojochuang):
Release note: Added
{{ozone.scm.container.placement.rack.scatter.capacity.aware.enabled}} (default
{{false}}). When enabled, the RackScatter EC placement policy makes intra-rack
datanode selection capacity-aware, picking the less utilized of two randomly
chosen datanodes so emptier nodes are preferred; disabled by default, so
existing behavior is unchanged.
> RackScatter policy does not consider node capacity
> --------------------------------------------------
>
> Key: HDDS-15093
> URL: https://issues.apache.org/jira/browse/HDDS-15093
> Project: Apache Ozone
> Issue Type: Bug
> Reporter: Wei-Chiu Chuang
> Assignee: Chi-Hsuan Huang
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.3.0
>
>
> EC container placement adopts rack-scatter policy where stripes are placed in
> different racks.
>
> However, the datanode within the same rack is chosen at random. Let's say a
> rack has both 100TB datanode and 400TB datanode, the 100TB datanode will fill
> up sooner than the 400TB datanode. There is no capacity consideration
> implemented intra-rack.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
