[infinispan-issues] [JBoss JIRA] (ISPN-11641) JVM crashed. SIGSEGV in org.rocksdb.RocksIterator.next0

[Dan Berindei (Jira)]

    [ https://issues.redhat.com/browse/ISPN-11641?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14168450#comment-14168450 ] 

Dan Berindei commented on ISPN-11641:
-------------------------------------

Looks like [https://github.com/facebook/rocksdb/issues/5234]

[~william.burns] could we somehow cancel all in-progress cache iterations/publishers as the first step of cache.stop()? Or maybe there's something simpler we can do just for RocksDBStore, since the other stores can catch any exception an replace it with an {{IllegalLifecycleStateException}}?

> JVM crashed. SIGSEGV in org.rocksdb.RocksIterator.next0
> -------------------------------------------------------
>
>                 Key: ISPN-11641
>                 URL: https://issues.redhat.com/browse/ISPN-11641
>             Project: Infinispan
>          Issue Type: Bug
>          Components: Loaders and Stores
>    Affects Versions: 9.4.16.Final
>            Reporter: Sergey Morgunov
>            Priority: Major
>         Attachments: crash.log
>
>
> *JRE*: OpenJDK 8u222.
> Crash log attached. (crash.log)
> Cache configuration:
> {code:xml}
> <distributed-cache name="sessions" configuration="sessionsConfiguration">
>     <memory>
>         <object size="100000" strategy="REMOVE"/>
>     </memory>
>     <persistence passivation="false">
>         <rocksdb-store path="cache/warm/" preload="true" fetch-state="true">
>             <expiration path="cache/expired/"/>
>         </rocksdb-store>
>     </persistence>
> </distributed-cache>
> {code}



--
This message was sent by Atlassian Jira
(v7.13.8#713008)