[
https://issues.apache.org/jira/browse/HDFS-12979?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16834199#comment-16834199
]
Chen Liang commented on HDFS-12979:
-----------------------------------
Post v003 patch which uses the alternative approach of having more than one
primary checkpointers. Namely, different ANN/Observer can have different SbN as
their primary.
Upside is that no need to add the new http code, and the image uploading
workload gets distributed across SbNs; Downside is that it seems ANN/Observer
could have different images on their disk because of different primary check
pointers, so when system restart, there is a possibility that they may pick up
different images then ANN/Observer could potentially run into different txids.
Which may or may not be a problem. Still need to verify this, if yes, we need
to be careful about this.
Any thoughts [~shv]?
> StandbyNode should upload FsImage to ObserverNode after checkpointing.
> ----------------------------------------------------------------------
>
> Key: HDFS-12979
> URL: https://issues.apache.org/jira/browse/HDFS-12979
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: hdfs
> Reporter: Konstantin Shvachko
> Assignee: Chen Liang
> Priority: Major
> Attachments: HDFS-12979.001.patch, HDFS-12979.002.patch,
> HDFS-12979.003.patch
>
>
> ObserverNode does not create checkpoints. So it's fsimage file can get very
> old making bootstrap of ObserverNode too long. A StandbyNode should copy
> latest fsimage to ObserverNode(s) along with ANN.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
