[
https://issues.apache.org/jira/browse/FLINK-34502?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Rui Fan updated FLINK-34502:
----------------------------
Description:
This is follow up Jira of FLINK-34471.
FLINK-34471 assuming all connections type are ALL_TO_ALL. This Jira will
optimize it to save some network memory for forward and rescale connection.
Following is the optimization logic:
* When the connection type is FORWARD and the parallelism of 2 tasks are same.
We calculate network memory based on FORWARD connection.
* When the connection type is FORWARD and the parallelism of 2 tasks are not
same. We calculate network memory based on RESCALE connection.
* When the connection type is RESCALE. We calculate network memory based on
RESCALE connection.
was:
This is follow up Jira of FLINK-34471.
FLINK-34471 assuming all connections type are ALL_TO_ALL. This Jira will
optimize it to save some network memory for forward and rescale connection.
> Support calculating network memory for forward and rescale edge
> ---------------------------------------------------------------
>
> Key: FLINK-34502
> URL: https://issues.apache.org/jira/browse/FLINK-34502
> Project: Flink
> Issue Type: Improvement
> Components: Autoscaler
> Reporter: Rui Fan
> Assignee: Rui Fan
> Priority: Major
>
> This is follow up Jira of FLINK-34471.
> FLINK-34471 assuming all connections type are ALL_TO_ALL. This Jira will
> optimize it to save some network memory for forward and rescale connection.
> Following is the optimization logic:
> * When the connection type is FORWARD and the parallelism of 2 tasks are
> same. We calculate network memory based on FORWARD connection.
> * When the connection type is FORWARD and the parallelism of 2 tasks are not
> same. We calculate network memory based on RESCALE connection.
> * When the connection type is RESCALE. We calculate network memory based on
> RESCALE connection.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
