curator-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Germán Blanco (JIRA) <>
Subject [jira] [Commented] (CURATOR-47) Two leaders in the leader election recipe
Date Wed, 17 Jul 2013 13:22:48 GMT


Germán Blanco commented on CURATOR-47:


thanks again for looking at the test. There did seem to be a few things wrong with it. However,
after solving all the things I found, something that might be a problem is still there. I
have modified your test with only three changes:
- Adding a constant NO_ONE = -1 to set the leader when there is no leader (you were using
0 which was also the value for one of the participants.
- Changing session timeout and connection time out for the Curator client.
- revers the order of the log and semaphore release for the connection lost/suspended event
(this might not be required).

Here is the result:

With this changes I got the "More than one leader" message using your test. Please, help me
find why.
Are there any restrictions on the session and connection timeouts?
> Two leaders in the leader election recipe
> -----------------------------------------
>                 Key: CURATOR-47
>                 URL:
>             Project: Apache Curator
>          Issue Type: Bug
>          Components: Recipes
>    Affects Versions: 2.1.0-incubating
>         Environment: Windows and Java 1.7
>            Reporter: Germán Blanco
>             Fix For: TBD
>         Attachments: patch_test_rolling.patch,
> When using the Leader Election recipe and doing a rolling restart of the zookeeper servers
in the cluster, it seems that at some points there are two participants that see themselves
as leaders.

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see:

View raw message