[
https://issues.jboss.org/browse/WFCORE-1912?page=com.atlassian.jira.plugi...
]
Brian Stansberry commented on WFCORE-1912:
------------------------------------------
The rollback behavior of the deployment scanner is configurable via the
runtime-failure-causes-rollback attribute and the standard setting in our config files is
'false'.
It's not practical to control this behavior from a descriptor, as the MSC services
involved in deployment do not have access to any mechanism to control the decisions of the
management layer about rollback. It's also not possible to control rollback based on
an individual resource's configuration. Rollback behavior is per overall request, and
a request can involve multiple resources. Boot basically implemented as a single request
and a composite op is a single request potentially involving multiple resources. An
individual resource should not be able to dictate the rollback behavior of a request that
involves other resources.
Redeploy deployment if all missing dependencies for deployment are
corrected
-----------------------------------------------------------------------------
Key: WFCORE-1912
URL:
https://issues.jboss.org/browse/WFCORE-1912
Project: WildFly Core
Issue Type: Enhancement
Components: Server
Reporter: Miroslav Novak
Assignee: Jason Greene
Labels: user_experience
If deployment failed to deploy and later all missing dependencies for deployment are
corrected then deployment would redeploy automatically. This behavior could be
configurable.
For example if you deploy EJB which injects queue like:
{code}
@Resource(mappedName = "java:/jms/queue/OutQueue")
private Queue outQueue;
{code}
and administrator deploys queue later then EJB would automatically get redeployed.
--
This message was sent by Atlassian JIRA
(v7.2.2#72004)