[
https://issues.apache.org/jira/browse/HBASE-22539?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16901083#comment-16901083
]
Sean Busbey commented on HBASE-22539:
-------------------------------------
{quote}
bq. Other possible options. It is too late for such a discussion.
Well, I had made that comment way before the PR was merged, but got ignored.
Still, if folks think it's worth it, I can work on a separate jira, and once we
all agree on an optimal solution, we revert the changes from this commit.
{quote}
Just from someone who's thus far only lurked on this particular issue, it's
almost never too late for discussion especially when we obviously don't have
consensus. It's fine to give particular concerns around different approaches.
It's even better when those concerns include a willingness to implement things
that address them (as both [~wchevreuil] and [~openinx] have done above).
{quote}
Ping Wellington Chevreuil, what's your opinion? I'm going to release 2.1.6
now...
{quote}
We have a community norm across ASF projects of accounting for asynchronous
communication on the order of across a 3 day window.
Unless I missed a VOTE over the weekend, it's a release candidate you have
queued up. If you find this error sufficiently severe to warrant the approach
of releasing a quick fix and following up later, that's your prerogative. (I
personally think there are sufficient workarounds available.) AFAICT this
doesn't impact the compatibility surface in any way. We're a do-acracy and you
can feel free to stage a 2.1.6 candidate now or after waiting to hear back from
Wellington. If he thinks the current band-aid has drawbacks severe enough to
warrant not including it even as a temporary measure, he can always give that
feedback on an RC thread.
> WAL corruption due to early DBBs re-use when Durability.ASYNC_WAL is used
> -------------------------------------------------------------------------
>
> Key: HBASE-22539
> URL: https://issues.apache.org/jira/browse/HBASE-22539
> Project: HBase
> Issue Type: Bug
> Components: rpc, wal
> Affects Versions: 2.2.0, 2.0.5, 2.1.5
> Reporter: Wellington Chevreuil
> Assignee: Duo Zhang
> Priority: Blocker
> Fix For: 3.0.0, 2.3.0, 2.0.6, 2.2.1, 2.1.6
>
> Attachments: HBASE-22539-UT.patch, HBASE-22539.branch-2.001.patch
>
>
> Summary
> We had been chasing a WAL corruption issue reported on one of our customers
> deployments running release 2.1.1 (CDH 6.1.0). After providing a custom
> modified jar with the extra sanity checks implemented by HBASE-21401 applied
> on some code points, plus additional debugging messages, we believe it is
> related to DirectByteBuffer usage, and Unsafe copy from offheap memory to
> on-heap array triggered
> [here|https://github.com/apache/hbase/blob/branch-2.1/hbase-common/src/main/java/org/apache/hadoop/hbase/util/ByteBufferUtils.java#L1157],
> such as when writing into a non ByteBufferWriter type, as done
> [here|https://github.com/apache/hbase/blob/branch-2.1/hbase-common/src/main/java/org/apache/hadoop/hbase/io/ByteBufferWriterOutputStream.java#L84].
> More details on the following comment.
>
--
This message was sent by Atlassian JIRA
(v7.6.14#76016)
