[
https://issues.apache.org/jira/browse/IGNITE-27118?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Aleksandr Chesnokov reassigned IGNITE-27118:
--------------------------------------------
Assignee: Aleksandr Chesnokov
> Cannot disable rolling upgrade
> ------------------------------
>
> Key: IGNITE-27118
> URL: https://issues.apache.org/jira/browse/IGNITE-27118
> Project: Ignite
> Issue Type: Bug
> Reporter: Aleksandr Chesnokov
> Assignee: Aleksandr Chesnokov
> Priority: Major
> Labels: IEP-132, ise
> Attachments: IGNITE-27118_Test_fail.patch
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> There is a scenario when we cannot disable rolling upgrade even if there is
> only one node in cluster
> Steps to reproduce:
> # Set TcpDiscoverySpi#setJoinTimeout to zero (or very big number)
> # Start server grid v2.18.0. It is coordinator now
> # Enable rolling upgrade with targetVersion=2.18.1
> # Start another grid v2.18.1 and stop it
> # Try to disable rolling upgrade on coordinator. It fails with:
> {noformat}
> Can't disable rolling upgrade with different versions in cluster:
> 2.18.0#20251120-sha1:d0e84887, 2.18.1#19700101-sha1:00000000
> {noformat}
>
> Reproducer is attached. The problem is related to
> org/apache/ignite/internal/processors/rollingupgrade/RollingUpgradeProcessor.java:235
> New node left ring but the processor thinks that new node is still joining
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
