[ 
https://issues.apache.org/jira/browse/CASSANDRA-15214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17463991#comment-17463991
 ] 

Yifan Cai edited comment on CASSANDRA-15214 at 12/22/21, 5:59 PM:
------------------------------------------------------------------

[~paulo], your suggestion makes sense. The synthetic heap OOM assumes users run 
cassandra with the on OOM flags. So it is reasonable to check the existence of 
the related options. I started a separate ticket: CASSANDRA-17227. 
Would the JVM result into a worse state due to the synthetic heap OOM? Probably 
not. The (actual) OOM happened and the JVM is in a bad state already. 


was (Author: yifanc):
[~paulo], your suggestion makes sense. The synthetic heap OOM assumes users run 
cassandra with the on OOM flags. So it is reasonable to check the existence of 
the related options. I will start a separate ticket. 
Would the JVM result into a worse state due to the synthetic heap OOM? Probably 
not. The (actual) OOM happened and the JVM is in a bad state already. 

> Internode messaging catches OOMs and does not rethrow
> -----------------------------------------------------
>
>                 Key: CASSANDRA-15214
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-15214
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Messaging/Client, Messaging/Internode
>            Reporter: Benedict Elliott Smith
>            Assignee: Yifan Cai
>            Priority: Normal
>             Fix For: 4.0-beta4, 4.0
>
>         Attachments: oom-experiments.zip
>
>          Time Spent: 1h 40m
>  Remaining Estimate: 0h
>
> Netty (at least, and perhaps elsewhere in Executors) catches all exceptions, 
> so presently there is no way to ensure that an OOM reaches the JVM handler to 
> trigger a crash/heapdump.
> It may be that the simplest most consistent way to do this would be to have a 
> single thread spawned at startup that waits for any exceptions we must 
> propagate to the Runtime.
> We could probably submit a patch upstream to Netty, but for a guaranteed 
> future proof approach, it may be worth paying the cost of a single thread.



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org

Reply via email to