[
https://issues.jboss.org/browse/JBIDE-21863?page=com.atlassian.jira.plugi...
]
Viacheslav Kabanovich edited comment on JBIDE-21863 at 3/14/16 4:50 PM:
------------------------------------------------------------------------
When WatchManager.stopWatch is called in the context of
DeploymentResourceMapper.buildDeployments(), it seems buildDeployments() does not want
that watch for this project be restarted until it explicitly calls in finally block
WatchManager.getInstance().startWatch(project). So, maybe in this context a flag may be
set on WatchManager and calls to WatchListener for the project may be just ignored until
buildDeployments() explicitly restarts the watch and remove the flag?
was (Author: scabanovich):
When WatchManager.stopWatch is called in the context of
DeploymentResourceMapper.buildDeployments(), it seems buildDeployments(project) does not
want that watch for this project be restarted until it explicitly calls in finally block
WatchManager.getInstance().startWatch(project). So, maybe in this context a flag may be
set on WatchManager and calls to WatchListener for the project may be just ignored until
buildDeployments() explicitly restarts the watch and remove the flag?
WatchManager makes multiple callbacks for same event
----------------------------------------------------
Key: JBIDE-21863
URL:
https://issues.jboss.org/browse/JBIDE-21863
Project: Tools (JBoss Tools)
Issue Type: Bug
Components: openshift
Affects Versions: 4.3.1.Beta1
Reporter: Jeff Cantrill
Assignee: Jeff Cantrill
Priority: Blocker
Fix For: 4.4.0.Alpha1
There is an issue in either the WatchClient or the the dependent functionality of the the
openshift-rest-client where a new listener is registered when the watch restarts from a
timeout. Listeners should only be able registered once
--
This message was sent by Atlassian JIRA
(v6.4.11#64026)