[
https://issues.apache.org/jira/browse/FLINK-20468?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrey Zagrebin updated FLINK-20468:
------------------------------------
Description:
[Original
discussion|https://github.com/apache/flink/pull/14199#issuecomment-733659726]
Currently, there is no easy way to test how JM failover (revoke and grant
leadership) affects other features with the MiniCluster and its testing
resource rule. The custom HA services can be provided to the TestingMiniCluster
but there is no simple HA services to support revoking and granting leadership
with a valid in-memory checkpoint store. Providing a way to enable such
embedded HA services for the MiniCluster out of the box allows to implement IT
cases similar to E2E tests.
> Enable leadership control in MiniCluster to test JM failover
> ------------------------------------------------------------
>
> Key: FLINK-20468
> URL: https://issues.apache.org/jira/browse/FLINK-20468
> Project: Flink
> Issue Type: Improvement
> Components: Runtime / Coordination
> Reporter: Andrey Zagrebin
> Assignee: Andrey Zagrebin
> Priority: Major
> Fix For: 1.12.0
>
>
> [Original
> discussion|https://github.com/apache/flink/pull/14199#issuecomment-733659726]
> Currently, there is no easy way to test how JM failover (revoke and grant
> leadership) affects other features with the MiniCluster and its testing
> resource rule. The custom HA services can be provided to the
> TestingMiniCluster but there is no simple HA services to support revoking and
> granting leadership with a valid in-memory checkpoint store. Providing a way
> to enable such embedded HA services for the MiniCluster out of the box allows
> to implement IT cases similar to E2E tests.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
