[
https://issues.apache.org/jira/browse/IGNITE-11798?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17021922#comment-17021922
]
Andrey Mashenkov commented on IGNITE-11798:
-------------------------------------------
[~tledkov-gridgain],
Looks like PartitionsExchangeAware.onDoneAfterTopologyUnlock() is called on
every topology change event.
So, mass client join\leave topology can start multiple partition reservation
invalidation tasks for same affinity version, isn't it?
> Memory leak on unstable topology caused by partition reservation
> ----------------------------------------------------------------
>
> Key: IGNITE-11798
> URL: https://issues.apache.org/jira/browse/IGNITE-11798
> Project: Ignite
> Issue Type: Bug
> Components: cache, sql
> Affects Versions: 2.7
> Reporter: Pavel Vinokurov
> Assignee: Taras Ledkov
> Priority: Major
> Attachments: PartitionReservationReproducer.java
>
> Time Spent: 10m
> Remaining Estimate: 0h
>
> Executing queries on unstable topology leads to OOM caused by leak of the
> partition reservation.
> The reproducer is attached.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
