[
https://issues.apache.org/jira/browse/HDFS-6440?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14240366#comment-14240366
]
Lei (Eddy) Xu commented on HDFS-6440:
-------------------------------------
[~jesse_yates] Thanks for working on this cool feature. We have read your
design doc and came up only a few questions:
# What is the procedure for adding or replacing NNs? Could it support
dynamically adding NNs without downtime?
# It seems that whether to upload a fsimage is mostly determined by SNN (e.g.,
finishing a checkpoint). Would it be possible to avoid mulitple SNNs to upload
fsimages with trivial deltas in a short time? E.g., let ANN to reject upload
requests if {{lastUploadTime > now - quiet period && num of edits < N}} ?
# It seems that QJM inherits the behaviors from the current ANN/SNN design that
it will purge edit logs after *_one_* SNN uploads a fsimage. Would it be
possible that this behavior makes other SNNs miss the edit logs? E.g., if a SNN
crashes and comes back online, but the edit logs are purged?
# Does this work support rolling upgrade?
# Would it makes client failover more complicated?
And some minor concerns:
# What would be the impact on the DN side?
# What are the changes on the test resources files (hadoop-*-reserved.tgz) ?
Thanks again for this awesome work!
> Support more than 2 NameNodes
> -----------------------------
>
> Key: HDFS-6440
> URL: https://issues.apache.org/jira/browse/HDFS-6440
> Project: Hadoop HDFS
> Issue Type: New Feature
> Components: auto-failover, ha, namenode
> Affects Versions: 2.4.0
> Reporter: Jesse Yates
> Assignee: Jesse Yates
> Attachments: Multiple-Standby-NameNodes_V1.pdf,
> hdfs-6440-cdh-4.5-full.patch, hdfs-multiple-snn-trunk-v0.patch
>
>
> Most of the work is already done to support more than 2 NameNodes (one
> active, one standby). This would be the last bit to support running multiple
> _standby_ NameNodes; one of the standbys should be available for fail-over.
> Mostly, this is a matter of updating how we parse configurations, some
> complexity around managing the checkpointing, and updating a whole lot of
> tests.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
