[
https://issues.apache.org/jira/browse/HDFS-3519?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ming Ma updated HDFS-3519:
--------------------------
Attachment: HDFS-3519.patch
To follow up on this, https://issues.apache.org/jira/browse/HDFS-4811 discussed
about adding timestamp to checkpoint file name to manage the concurrency.
Alternatively, we can move {{ImageServlet}}'s code that handle s duplicated
download requests to {{FSImage}} so that it covers regular checkpoint. Any
thoughts on this?
In addition, {{currentlyDownloadingCheckpoints}} is declared as static. That is
problematic in unit test where two namenodes share the same list. It doesn't
need to be static as {{ImageServlet}} gets the global FSImage object from
{{NameNodeHttpServer}}.
Another thing about the test case, it is possible that the first standby nn
checked by the test code finishes the checkpoint and upload before the second
standby nn kicks off its checkpoint process; it doesn't appears to be the
intention of the test case. We can increase {{SlowCodec}} sleep interval for
that.
Chris, sorry didn't checkout the jira owner until after I write up the patch.
> Checkpoint upload may interfere with a concurrent saveNamespace
> ---------------------------------------------------------------
>
> Key: HDFS-3519
> URL: https://issues.apache.org/jira/browse/HDFS-3519
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Affects Versions: 1.0.3, 2.0.0-alpha
> Reporter: Todd Lipcon
> Assignee: Chris Nauroth
> Priority: Critical
> Attachments: HDFS-3519.patch, test-output.txt
>
>
> TestStandbyCheckpoints failed in [precommit build
> 2620|https://builds.apache.org/job/PreCommit-HDFS-Build/2620//testReport/]
> due to the following issue:
> - both nodes were in Standby state, and configured to checkpoint "as fast as
> possible"
> - NN1 starts to save its own namespace
> - NN2 starts to upload a checkpoint for the same txid. So, both threads are
> writing to the same file fsimage.ckpt_12, but the actual file contents
> correspond to the uploading thread's data.
> - NN1 finished its saveNamespace operation while NN2 was still uploading. So,
> it renamed the ckpt file. However, the contents of the file are still empty
> since NN2 hasn't sent any bytes
> - NN2 finishes the upload, and the rename() call fails, which causes the
> directory to be marked failed, etc.
> The result is that there is a file fsimage_12 which appears to be a finalized
> image but in fact is incompletely transferred. When the transfer completes,
> the problem "heals itself" so there wouldn't be persistent corruption unless
> the machine crashes at the same time. And even then, we'd still have the
> earlier checkpoint to restore from.
> This same race could occur in a non-HA setup if a user puts the NN in safe
> mode and issues saveNamespace operations concurrent with a 2NN checkpointing,
> I believe.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
