You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@cassandra.apache.org by "Eric Evans (JIRA)" <ji...@apache.org> on 2009/07/08 19:11:14 UTC
[jira] Created: (CASSANDRA-282) system tests fail straingely if an
instance is running
system tests fail straingely if an instance is running
------------------------------------------------------
Key: CASSANDRA-282
URL: https://issues.apache.org/jira/browse/CASSANDRA-282
Project: Cassandra
Issue Type: Bug
Reporter: Eric Evans
Assignee: Eric Evans
The systems tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start it's instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-282) system tests fail straingely if
an instance is running
Posted by "Hudson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12729623#action_12729623 ]
Hudson commented on CASSANDRA-282:
----------------------------------
Integrated in Cassandra #133 (See [http://hudson.zones.apache.org/hudson/job/Cassandra/133/])
abort system tests if previous run shutdown uncleanly
Patch by eevans; reviewed by Michael Greene for
move ports used in system tests; shorten timeout
Patch by Michael Greene; reviewed by eevans for
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Michael Greene
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-282) system tests fail straingely if an
instance is running
Posted by "Eric Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Eric Evans updated CASSANDRA-282:
---------------------------------
Description: The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages. (was: The systems tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start it's instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left from failed tests, and at least produce clearer error messages.)
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Eric Evans
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-282) system tests fail straingely if
an instance is running
Posted by "Eric Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12729331#action_12729331 ]
Eric Evans commented on CASSANDRA-282:
--------------------------------------
+1 282-v1.diff
0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt is meant to be applied in addition to Michael's (282-v1.diff), and causes the tests to fail with a (more )useful message if a previous run failed to cleanup.
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Eric Evans
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (CASSANDRA-282) system tests fail straingely if
an instance is running
Posted by "Michael Greene (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12729344#action_12729344 ]
Michael Greene commented on CASSANDRA-282:
------------------------------------------
+1 on 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt since it actually caught the error it was supposed to on my system, even though I didn't expect it.
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Eric Evans
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-282) system tests fail straingely if an
instance is running
Posted by "Michael Greene (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Michael Greene updated CASSANDRA-282:
-------------------------------------
Component/s: Tools
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Components: Tools
> Reporter: Eric Evans
> Assignee: Michael Greene
> Fix For: 0.4
>
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-282) system tests fail straingely if an
instance is running
Posted by "Eric Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Eric Evans updated CASSANDRA-282:
---------------------------------
Attachment: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Eric Evans
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (CASSANDRA-282) system tests fail straingely if an
instance is running
Posted by "Eric Evans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Eric Evans resolved CASSANDRA-282.
----------------------------------
Resolution: Fixed
Assignee: Michael Greene (was: Eric Evans)
committed.
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Michael Greene
> Attachments: 0001-CASSANDRA-282-stop-drop-roll-if-uncleanly-shutdown.txt, 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (CASSANDRA-282) system tests fail straingely if an
instance is running
Posted by "Michael Greene (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/CASSANDRA-282?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Michael Greene updated CASSANDRA-282:
-------------------------------------
Attachment: 282-v1.diff
This patch shifts all the relevant ports by 10, so they are still predictable. It also sets the timeout for a new client connection to 10 seconds, because 20 seconds gets really unreasonable if you are trying to debug things, and I can't imagine the regression test server taking longer than 10 seconds to come up.
> system tests fail straingely if an instance is running
> ------------------------------------------------------
>
> Key: CASSANDRA-282
> URL: https://issues.apache.org/jira/browse/CASSANDRA-282
> Project: Cassandra
> Issue Type: Bug
> Reporter: Eric Evans
> Assignee: Eric Evans
> Attachments: 282-v1.diff
>
>
> The system tests blow up spectacularly if there is another instance of cassandra running (and it can be difficult to tell that this is why). Having the harness start its instance using alternate ports would prevent this from happening in some cases, but something should be done to test for instances left over from failed tests, and at least produce clearer error messages.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.