[
https://issues.apache.org/jira/browse/HADOOP-18569?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17646090#comment-17646090
]
ASF GitHub Bot commented on HADOOP-18569:
-----------------------------------------
adoroszlai opened a new pull request, #5207:
URL: https://github.com/apache/hadoop/pull/5207
## What changes were proposed in this pull request?
After upgrading Netty from 4.1.68 to 4.1.77 (HADOOP-18079), NFS Gateway
started randomly crashing when writing data (can be easily reproduced by a few
10MB+ files). The problem was triggered by [reduced default chunk size in
PooledByteBufAllocator](https://github.com/netty/netty/commit/f650303911) (in
4.1.75), but it turned out to be caused by a buffer released too early in NFS
Gateway (HADOOP-11245).
https://issues.apache.org/jira/browse/HADOOP-18569
## How was this patch tested?
Deployed cluster with the change, tested write/read via NFS mount.
> NFS Gateway may release buffer too early
> ----------------------------------------
>
> Key: HADOOP-18569
> URL: https://issues.apache.org/jira/browse/HADOOP-18569
> Project: Hadoop Common
> Issue Type: Bug
> Components: nfs
> Affects Versions: 3.4.0, 3.3.5, 3.2.5, 3.3.9
> Reporter: Attila Doroszlai
> Assignee: Attila Doroszlai
> Priority: Blocker
>
> After upgrading Netty from 4.1.68 to 4.1.77 (HADOOP-18079), NFS Gateway
> started crashing when writing data (can be easily reproduced by a few 10MB+
> files). The problem was triggered by [reduced default chunk size in
> PooledByteBufAllocator|https://github.com/netty/netty/commit/f650303911] (in
> 4.1.75), but it turned out to be caused by a buffer released too early in NFS
> Gateway.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
