[
https://issues.apache.org/jira/browse/GEODE-8968?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17312493#comment-17312493
]
ASF subversion and git services commented on GEODE-8968:
--------------------------------------------------------
Commit 8935ccd0a64fd871a9d8c3d41afa408025a567aa in geode-native's branch
refs/heads/develop from Mario Salazar de Torres
[ https://gitbox.apache.org/repos/asf?p=geode-native.git;h=8935ccd ]
GEODE-8968: Fix toDataMutable coredump (#770)
- In those scenarios in which Pdx serialization occurs while
PdxTypeRegistry is being cleaned up, it happened that NC crashed
because it couldn't find the PdxType.
- This commit adds retries to Pdx serialization, and in case retries
are exhausted, UnknownPdxTypeException is thrown, so it can be
handled by the user.
- Implemented IT to verify this behaviour.
> PdxTypeRegistry clean up causes a coredump
> ------------------------------------------
>
> Key: GEODE-8968
> URL: https://issues.apache.org/jira/browse/GEODE-8968
> Project: Geode
> Issue Type: Bug
> Components: native client
> Affects Versions: 1.14.0
> Reporter: Mario Salazar de Torres
> Priority: Major
> Labels: pull-request-available
>
> *GIVEN* a client/server that with PDX serialization
> *WHEN* the client looses connection towards the cluster
> *AND* PdxTypeRegistry is cleaned up
> *WHILE* a PDX instance is being serialized
> *THEN* a coredump occurs
> ----
> *Additional information:*
> As it seems PdxHelper::serializePdx calls toData -> toDataMutable ->
> getPdxType and the later part returns a nullptr if the registry was cleaned
> up in the process.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
