[
https://issues.apache.org/jira/browse/HDFS-6800?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14081743#comment-14081743
]
Colin Patrick McCabe commented on HDFS-6800:
--------------------------------------------
Just to give a bit of additional context here: [~james.thomas] did some
benchmarks that showed that "creating 200k hard links (100k blocks) take just
over a second on a dual core PC" when using the optimized hard link code. On a
performance basis, I think it's very feasible to support rolling upgrade to new
DN layout versions.
If we don't choose to support this, we are going to make it very hard to evolve
the DN code in the future. We would then require a major version change (i.e.
Hadoop 3.0) to make any major DN changes. So I think we should just change the
documentation a bit and support this in the obvious way... by having the users
call {{datanode \-rollback}} during a rolling rollback if needed. What do you
guys think?
> Detemine how Datanode layout changes should interact with rolling upgrade
> -------------------------------------------------------------------------
>
> Key: HDFS-6800
> URL: https://issues.apache.org/jira/browse/HDFS-6800
> Project: Hadoop HDFS
> Issue Type: Improvement
> Components: datanode
> Affects Versions: 2.6.0
> Reporter: Colin Patrick McCabe
>
> We need to handle attempts to rolling-upgrade the DataNode to a new storage
> directory layout.
> One approach is to disallow such upgrades. If we choose this approach, we
> should make sure that the system administrator gets a helpful error message
> and a clean failure when trying to use rolling upgrade to a version that
> doesn't support it. Based on the compatibility guarantees described in
> HDFS-5535, this would mean that *any* future DataNode layout changes would
> require a major version upgrade.
> Another approach would be to support rolling upgrade from an old DN storage
> layout to a new layout. This approach requires us to change our
> documentation to explain to users that they should supply the {{\-rollback}}
> command on the command-line when re-starting the DataNodes during rolling
> rollback. Currently the documentation just says to restart the DataNode
> normally.
> Another issue here is that the DataNode's usage message describes rollback
> options that no longer exist. The help text says that the DN supports
> {{\-rollingupgrade rollback}}, but this option was removed by HDFS-6005.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
