[
https://issues.apache.org/jira/browse/FLINK-37205?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dian Fu closed FLINK-37205.
---------------------------
Fix Version/s: 2.0.0
1.19.2
1.20.2
Resolution: Fixed
Fixed in:
- master via a498cf2c228680a44f71d81e7c08b39e1a968a1c
- release-2.0 via bfea5b52be951ca74e41bc066cf655f06f828376
- release-1.20 via 14e853ab832bd9aabc61219a9ce187c201e9603c
- release-1.19 via af041a2a883f31baf24537964e15e6679ba66dcd
> Correct the state cache behavior during bump beam version
> ---------------------------------------------------------
>
> Key: FLINK-37205
> URL: https://issues.apache.org/jira/browse/FLINK-37205
> Project: Flink
> Issue Type: Bug
> Components: API / Python
> Affects Versions: 1.17.0, 1.18.0, 1.17.1, 1.17.2, 1.19.0, 1.18.1, 1.20.0,
> 1.19.1
> Reporter: Dian Fu
> Assignee: Dian Fu
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.0.0, 1.19.2, 1.20.2
>
>
> It has changed the state cache strategy from size based to bytes based since
> Beam 2.42.0 (See https://github.com/apache/beam/pull/22924 for more details).
> We should also support memory based state cache in the long term. Before
> that, we should correct the behavior which seems broken after bumping the
> Beam version. It may cause the state cache continuously increase and finally
> cause the job OOM.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
