[
https://issues.apache.org/jira/browse/IGNITE-20424?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Mikhail Petrov updated IGNITE-20424:
------------------------------------
Description:
All messages designated for the thin client are accumulated in the selector
queue GridSelectorNioSessionImpl#queue before they are sent. Note that
GridSelectorNioSessionImpl#queue is unbound. If the thin client connection is
slow and a huge number of messages with a large weight will be sent to it,
GridSelectorNioSessionImpl#queue growing size can lead to OOM.
See
1. https://issues.apache.org/jira/browse/IGNITE-20327
2. https://github.com/apache/ignite/issues/10559
We need to investigate mechanisms to limit the thin client message queue size.
was:
All messages designated for the thin client are accumulated in the selector
queue GridSelectorNioSessionImpl#queue before they are sent. Note that
GridSelectorNioSessionImpl#queue is unbound. If the thin client connection is
slow and a huge number of messages with a large weight will be sent to it,
GridSelectorNioSessionImpl#queue growing size can lead to OOM.
See
https://issues.apache.org/jira/browse/IGNITE-20327 and
https://github.com/apache/ignite/issues/10559
for real world scenario to reproduce it.
We need to investigate mechanisms to limit the thin client message queue size.
> [Thin clients] Slow thin clients connection can lead to node failure with OOM
> -----------------------------------------------------------------------------
>
> Key: IGNITE-20424
> URL: https://issues.apache.org/jira/browse/IGNITE-20424
> Project: Ignite
> Issue Type: Task
> Reporter: Mikhail Petrov
> Priority: Major
>
> All messages designated for the thin client are accumulated in the selector
> queue GridSelectorNioSessionImpl#queue before they are sent. Note that
> GridSelectorNioSessionImpl#queue is unbound. If the thin client connection
> is slow and a huge number of messages with a large weight will be sent to it,
> GridSelectorNioSessionImpl#queue growing size can lead to OOM.
> See
> 1. https://issues.apache.org/jira/browse/IGNITE-20327
> 2. https://github.com/apache/ignite/issues/10559
> We need to investigate mechanisms to limit the thin client message queue size.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)