[
https://issues.apache.org/jira/browse/FLUME-2642?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14356190#comment-14356190
]
Hari Shreedharan edited comment on FLUME-2642 at 3/11/15 3:45 AM:
------------------------------------------------------------------
FYI, there is an edge case here -- where multiple threads could read the
{{lastReloginAttempt}}, which could be > 5 mins. Now all of them could end up
calling checkTGT..So every 5 mins there is a possibility that many threads
could try to enter the checkTGT.. method. It is fine (the code is still
thread-safe), but it is worth noting in the comments.
was (Author: hshreedharan):
FYI, there is an edge case here -- where multiple threads could read the
{code}lastReloginAttempt{code}, which could be > 5 mins. Now all of them could
end up calling checkTGT..So every 5 mins there is a possibility that many
threads could try to enter the checkTGT.. method. It is fine (the code is still
thread-safe), but it is worth noting in the comments.
> Limit kerberos relogin attempt
> -------------------------------
>
> Key: FLUME-2642
> URL: https://issues.apache.org/jira/browse/FLUME-2642
> Project: Flume
> Issue Type: Improvement
> Affects Versions: v1.6.0
> Reporter: Johny Rufus
> Assignee: Johny Rufus
> Fix For: v1.6.0
>
> Attachments: FLUME-2642.patch
>
>
> UGI executor in flume authentication module attempts relogin using
> UserGroupInformation's checkTGTAndReloginFromKeytab().
> Since this method is synchronized, it would be better to space out the
> relogin attempts
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
