[
https://issues.apache.org/jira/browse/HDFS-15404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17148160#comment-17148160
]
Chen Liang edited comment on HDFS-15404 at 6/29/20, 9:53 PM:
-------------------------------------------------------------
Upload v004 patch to fix {{TestDFSHAAdminMiniCluster}} which was expecting to
see {{target_}} environment variables, while with this change, the fencing on
source will be using {{source_}} variables. This brings in another question to
my mind, which is that there could be existing tooling that relies on
{{target_*}} on both source and dst of the fencing. This might not be the right
use of the variables, but if they do exist, they may break. Do we plan to
support such use cases is the question, any thoughts [~shv]?
{{TestRollingUpgrade}} is passing in my local run. The other test fails are
known flaky tests AFAIK.
was (Author: vagarychen):
Upload v004 patch to fix {{TestDFSHAAdminMiniCluster}} which was expecting to
see {{target_*}} environment variables, while with this change, the fencing on
source will be using {{source_*}} variables. This brings in another question to
my mind, which is that there could be existing tooling that relies on
{{target_*}} on both source and dst of the fencing. This might not be the right
use of the variables, but if they do exist, they may break. Do we plan to
support such use cases is the question, any thoughts [~shv]?
{{TestRollingUpgrade}} is passing in my local run. The other test fails are
known flaky tests AFAIK.
> ShellCommandFencer should expose info about source
> --------------------------------------------------
>
> Key: HDFS-15404
> URL: https://issues.apache.org/jira/browse/HDFS-15404
> Project: Hadoop HDFS
> Issue Type: Improvement
> Reporter: Chen Liang
> Assignee: Chen Liang
> Priority: Major
> Attachments: HDFS-15404.001.patch, HDFS-15404.002.patch,
> HDFS-15404.003.patch, HDFS-15404.004.patch
>
>
> Currently the HA fencing logic in ShellCommandFencer exposes environment
> variable about only the fencing target. i.e. the $target_* variables as
> mentioned in this [document
> page|https://hadoop.apache.org/docs/current/hadoop-project-dist/hadoop-hdfs/HDFSHighAvailabilityWithQJM.html]).
>
> But here only the fencing target variables are getting exposed. Sometimes it
> is useful to expose info about the fencing source node. One use case is would
> allow source and target node to identify themselves separately and run
> different commands/scripts.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
