qpid-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alex Rudyy (JIRA)" <j...@apache.org>
Subject [jira] [Resolved] (QPID-8036) [Broker-J] [BDB HA] If ReplicationEnvironment fails when creating a node for the first time, clean up any directory left behind by BDB
Date Wed, 27 Dec 2017 14:14:00 GMT

     [ https://issues.apache.org/jira/browse/QPID-8036?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Alex Rudyy resolved QPID-8036.
------------------------------
    Resolution: Cannot Reproduce

I could not reproduce the reported issue. When node cannot join the group for whatever reason,
the creation of BDB HA VHN fails and {{AbstractConfiguredObject.CreateExceptionHandler#handleException}}
is invoked. As part of the latter call, the method {{BDBHAVirtualHostNodeImp#deleteAsync()}}
is executed which in turns calls {{BDBHAVirtualHostNodeImp#doDelete}}. The latter calls {{BDBConfigurationStore#onDelete}}
to remove the directory structure created by BDB JE environment. {{BDBConfigurationStore#onDelete}}
is called for both successful and unsuccessful execution of {{BDBHAVirtualHostNodeImp#doDelete}}
.It seems that entire JE working folder should be wiped out on any failure to create a replica
BDB HA VHN for existing HA  group.

> [Broker-J] [BDB HA] If ReplicationEnvironment fails when creating a node for the first
time, clean up any directory left behind by BDB
> --------------------------------------------------------------------------------------------------------------------------------------
>
>                 Key: QPID-8036
>                 URL: https://issues.apache.org/jira/browse/QPID-8036
>             Project: Qpid
>          Issue Type: Bug
>          Components: Broker-J
>    Affects Versions: qpid-java-6.0, qpid-java-6.1, qpid-java-broker-7.0.0
>            Reporter: Keith Wall
>             Fix For: qpid-java-broker-7.0.1
>
>
> If the {{ReplicatedEnvironment}} constructor fails for a new node, any directory left
behind by BDB JE HA should be cleaned up.  We have seen problem reports where group creation
had failed (owning to network issues within a segmented network), but the environment directory
was being left behind on disk.
> The existence of the directory then goes on to cause further failures if later another
group tries to create a new node in the same location.  Currently this requires manual (shell)
intervention to clean-up.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

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


Mime
View raw message