qpid-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF subversion and git services (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (QPID-6111) Ensure that ReplicatedEnvironmentFacade awaits sufficient time on shutdown for any in-progress restart to complete
Date Wed, 24 Sep 2014 13:03:34 GMT

    [ https://issues.apache.org/jira/browse/QPID-6111?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14146287#comment-14146287
] 

ASF subversion and git services commented on QPID-6111:
-------------------------------------------------------

Commit 1627305 from [~k-wall] in branch 'qpid/trunk'
[ https://svn.apache.org/r1627305 ]

QPID-6111: [Java Broker] Ensure that when the REF is shutdown sufficient time is allowed -
bug fix used Math.min rather than Math.max.

Also:
* reduced the default envSetupTimeout from 15mins to 3mins.
* log a warning if environment re-creation takes longer than 25% of the allowed timeout
* whilst restarting, recheck the REF state to avoid needlessly recreating the environment
if restart is no longer required (most likely owning to close)

> Ensure that ReplicatedEnvironmentFacade awaits sufficient time on shutdown for any in-progress
restart to complete 
> -------------------------------------------------------------------------------------------------------------------
>
>                 Key: QPID-6111
>                 URL: https://issues.apache.org/jira/browse/QPID-6111
>             Project: Qpid
>          Issue Type: Bug
>          Components: Java Broker
>    Affects Versions: 0.30
>            Reporter: Keith Wall
>            Assignee: Alex Rudyy
>             Fix For: 0.31
>
>         Attachments: 0001-QPID-6111-Java-Broker-HA-Ensure-that-when-the-REF-is.patch,
0002-QPID-6111-Java-Broker-HA-Tests-change-JE-tests-to-as.patch
>
>
> The HA system tests sporadically fail with exception "BindException: Address already
in use"  when trying to create a ReplicatedEnvironment.
> Analysis shows that this happens when a previous HA test has left behind an executor
(REF#_environmentJobExecutor) which is in the process of restarting the environment from a
previous test,   This leads to unexpected sporadic test failures.
> REF should be changed so that on close, it awaits any in progress restart.



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

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@qpid.apache.org
For additional commands, e-mail: dev-help@qpid.apache.org


Mime
View raw message