[
https://issues.apache.org/jira/browse/QPID-7560?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15751157#comment-15751157
]
Keith Wall edited comment on QPID-7560 at 12/15/16 1:49 PM:
------------------------------------------------------------
At the moment, the only thing that causes a VH to go into {{ERROR}} is a
failure to schedule the housekeeping task. This can only happen if the
housekeeping parameters are somehow misconfigured (for instance a negative
{{period}}). As you probably would not expect such parameters to be altered
in a production environment, I think the likelihood of a VH ending up in ERROR
to be small. If a VH were to somehow end up in ERROR, and an operator
attempt a restart, currently the Broker would fail if the {{onRestart}} route
were to be followed as the VH's CO would be duplicated. A restart would be
required to clear the condition.
Note: failure to recover the messages currently causes the Broker to go down
rather than go into ERROR state.
I have also checked the other COs and don't see other instances of this type of
problem.
was (Author: k-wall):
At the moment, the only thing that causes a VH to go into {{ERROR}} is a
failure to schedule the housekeeping task. This can only happen if the
housekeeping parameters are somehow misconfigured (for instance a negative
{{period}}). As you probably would not expect such parameters to be altered
in a production environment, I think the likelihood of a VH ending up in ERROR
to be small. If a VH were to somehow end up in ERROR, and an operator
attempt a restart, currently the Broker would fail if the {{onRestart}} route
were to be followed.
Note: failure to recover the messages currently causes the Broker to go down
rather than go into ERROR state.
I have also checked the other COs and don't see other instances of this type of
problem.
> AbstractVirtualHost defines two state transitions from ERROR to ACTIVE
> ----------------------------------------------------------------------
>
> Key: QPID-7560
> URL: https://issues.apache.org/jira/browse/QPID-7560
> Project: Qpid
> Issue Type: Bug
> Components: Java Broker
> Reporter: Keith Wall
> Fix For: qpid-java-6.2, qpid-java-6.1.1, qpid-java-6.0.6
>
>
> The current implementation of AbstractVirtualHost defines two state
> transitions from ERROR=>ACTIVE. The state change mechanics could choose
> either which could lead to undesired results.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@qpid.apache.org
For additional commands, e-mail: dev-h...@qpid.apache.org
