nifi-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mark Payne (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (NIFI-929) Ability to generate a 'true' pid file
Date Fri, 23 Oct 2015 13:01:27 GMT

    [ https://issues.apache.org/jira/browse/NIFI-929?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14970927#comment-14970927
] 

Mark Payne commented on NIFI-929:
---------------------------------

[~vs186031] - agreed. Both are responsible solutions. I would certainly opt for A) though.
I think in either case, there will still be a "pid" file and a "status" file. A) keeps backward
compatibility and in addition means that it's easier to obtain all of the values that we care
about by simply reading a single file. This way, the "status file" can be used by the bootstrap
and the .pid file is simply written out as a "useful tool" for other processes, such as Ambari.

> Ability to generate a 'true' pid file
> -------------------------------------
>
>                 Key: NIFI-929
>                 URL: https://issues.apache.org/jira/browse/NIFI-929
>             Project: Apache NiFi
>          Issue Type: Improvement
>          Components: Configuration
>    Affects Versions: 0.3.0
>            Reporter: Ali Bajwa
>            Priority: Minor
>
> The nifi pid file does not seem to be a true pid file (it has other info as well). 
> For integration with monitoring tools like Ambari it would be nice to have a true pid
file (containing only the pid) created as well (in line with other Hadoop components). This
is coming from the Ambari service I put together: I had to parse out the pid from Nifi pidfile
and maintain another one



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message