You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Gary Dusbabek (JIRA)" <ji...@apache.org> on 2010/01/13 20:36:54 UTC
[jira] Created: (CASSANDRA-695) Gossiper convicts same node over
and over.
Gossiper convicts same node over and over.
------------------------------------------
Key: CASSANDRA-695
URL: https://issues.apache.org/jira/browse/CASSANDRA-695
Project: Cassandra
Issue Type: Bug
Components: Core
Affects Versions: 0.5
Reporter: Gary Dusbabek
Assignee: Jaakko Laine
Priority: Minor
I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
1. Bring up several nodes.
2. Kill one of them.
Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-695:
-------------------------------------
Assignee: Gary Dusbabek (was: Jaakko Laine)
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Gary Dusbabek
> Priority: Trivial
> Fix For: 0.6
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12803512#action_12803512 ]
Jonathan Ellis commented on CASSANDRA-695:
------------------------------------------
+1
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Gary Dusbabek
> Priority: Trivial
> Fix For: 0.6
>
> Attachments: 695.patch
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jonathan Ellis updated CASSANDRA-695:
-------------------------------------
Fix Version/s: 0.6
Priority: Trivial (was: Minor)
Affects Version/s: (was: 0.5)
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Jaakko Laine
> Priority: Trivial
> Fix For: 0.6
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Issue Comment Edited: (CASSANDRA-695) Gossiper convicts same
node over and over.
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12803451#action_12803451 ]
Jonathan Ellis edited comment on CASSANDRA-695 at 1/21/10 7:46 PM:
-------------------------------------------------------------------
>From my reading this is Works As Intended. Lowering level to trace is fine with me. Go ahead and commit that, Gary.
was (Author: jbellis):
From my reading this is Works As Intended. Lowering level to trace is fine with me. Jaakko?
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Jaakko Laine
> Priority: Trivial
> Fix For: 0.6
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12803681#action_12803681 ]
Hudson commented on CASSANDRA-695:
----------------------------------
Integrated in Cassandra #331 (See [http://hudson.zones.apache.org/hudson/job/Cassandra/331/])
lower MS.convict debug statement to trace.
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Gary Dusbabek
> Priority: Trivial
> Fix For: 0.6
>
> Attachments: 695.patch
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Jonathan Ellis (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12803451#action_12803451 ]
Jonathan Ellis commented on CASSANDRA-695:
------------------------------------------
>From my reading this is Works As Intended. Lowering level to trace is fine with me. Jaakko?
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Jaakko Laine
> Priority: Minor
> Fix For: 0.6
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Gary Dusbabek (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gary Dusbabek updated CASSANDRA-695:
------------------------------------
Attachment: 695.patch
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Gary Dusbabek
> Priority: Trivial
> Fix For: 0.6
>
> Attachments: 695.patch
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (CASSANDRA-695) Gossiper convicts same node over
and over.
Posted by "Gary Dusbabek (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-695?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gary Dusbabek resolved CASSANDRA-695.
-------------------------------------
Resolution: Fixed
901893. Patch by Gary Dusbabek.
> Gossiper convicts same node over and over.
> ------------------------------------------
>
> Key: CASSANDRA-695
> URL: https://issues.apache.org/jira/browse/CASSANDRA-695
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Reporter: Gary Dusbabek
> Assignee: Gary Dusbabek
> Priority: Trivial
> Fix For: 0.6
>
> Attachments: 695.patch
>
>
> I can reproduce this in trunk or 0.5. It is quite easy to see with logging set to DEBUG.
> 1. Bring up several nodes.
> 2. Kill one of them.
> Gossip still continues to examine the failed node because it is in the ring, but gets convicted over and over at every check (see FailureDetector.interpret). If this is expected, we should consider lowering the debug statement in MessagingService.convict to TRACE.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.