[
https://issues.apache.org/jira/browse/HDFS-13767?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16567299#comment-16567299
]
Chen Liang commented on HDFS-13767:
-----------------------------------
Thanks [~shv] and [~xkrogen] for the review and the comments! Post v002 patch.
bq. It seems it would be better if there was a way for the AlignmentContext, a
generic/pluggable component, to make the decision about how to handle the call
I agree that this very HDFS-specific. An issue here is that while {{Server}}
has AlignmentContext instance, AlignmentContext has both server side and client
side implementation. And it can not be casted to GlobalStateIdContext because
it is in a different package. So adding {{shouldDeferProcessing(Server.Call
call)}}, or any method that checks whether to defer a call, means it will need
to be added in client alignment context as well. Which makes no sense to me.
The more I look at AlignmentContext, the more I think that it would be better
if we separate server and client AlignmentContext in some way. Then we can
introduce this defer check call to server side only. For now, I would prefer to
leave it like this.
bq. Why did you move setUpCluster() out of TestObserverNode#setUp()
When it was in {{setUp()}}, it is still effectively being called for each test.
Then, for the tests such as testMultiObserver, it gets called again in the test
method with a different parameter. So for these tests, setUpCluster() was being
called twice. This was a bug there. So I moved it to each test to fix this,
since {{setUpCluster}} takes a parameter that can be different for each test.
The other comments are addressed. I thought about using fast-tailing but was
being lazy to figure out how to enable it. Thanks a lot for the sharing Erik :)
! it does make the tests a lot faster.
> Add msync server implementation.
> --------------------------------
>
> Key: HDFS-13767
> URL: https://issues.apache.org/jira/browse/HDFS-13767
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: namenode
> Reporter: Chen Liang
> Assignee: Chen Liang
> Priority: Major
> Attachments: HDFS-13767-HDFS-12943.001.patch,
> HDFS-13767-HDFS-12943.002.patch, HDFS-13767.WIP.001.patch,
> HDFS-13767.WIP.002.patch, HDFS-13767.WIP.003.patch, HDFS-13767.WIP.004.patch
>
>
> This is a followup on HDFS-13688, where msync API is introduced to
> {{ClientProtocol}} but the server side implementation is missing. This is
> Jira is to implement the server side logic.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
