[
https://issues.jboss.org/browse/JBIDE-18862?page=com.atlassian.jira.plugi...
]
Max Rydahl Andersen commented on JBIDE-18862:
---------------------------------------------
Darryll - we already do all the stuff you are saying...what we have now is a bug in this
implementation that was missing the case where .dodeploy was possible to be present if you
don't start the server before first publish occurs.
That will be fixed; and that is what Rob is suggesting he fix. Btw. rob - we should fix
this in 4.2.x too.
To comment on Daryll steps:
{{{ Step 1) Eclipse instructs the AS to inhibit all deployment operations! }}}
Since AS7/WildFly do *not* start deploying things before there actually is a .dodeploy or
the last byte of an archive file is available this is already in place (sans the missing
removal of .dodeploy in the special case)
For pre-AS7 and the cases where jmx/api is enabled/availble we tell the server to stop
autoscanning for deployments.
{{{
Step 2) The AS positively acknowledges this instruction from Step 1 above (in the case of
a deployment is in progress at the AS side, it would not need to be aborted, it would
continue until success or failure, at that point the AS would return the positive
acknowledgement that all AS deployment operations have been temporary inhibited at
Eclipse's request).
}}}
If we had to wait for this smallest updates would be slow. There is no need for this when
Step 1 is done properly - which it is.
{{{
Step 3) Eclipse does publish operation (modifies files/dirs).
}}}
Yup.
{{{
Step 4) Eclipse instructs the AS the remove the inhibit of all deployment operations.
Ideally this might be done with remoting/EJBclient/HTTPwebsocket so if the socket
connection breaks/closes that also removes the inhibiter at the AS side.
}}}
I do not know what this means. But my guess is you mean the server should not start
deploying before we are done copying.
This happens today via .dodeploy marker or by using management operations or by the
archive being fully copied.
It is not an inhibitor it is an *activator*.
{{{
Step 5) This also allows the AS to immediately deploy what is sees. No waiting around a
few seconds for something to notice. Although I've not seen this problem in a while.
}}}
Yes, because that is what AS7/WF8 the jboss tools do. Note, if you use <Java7 there is
a 5 second delay by default to not slow other things down, >Java7 native java file
monitoring hooks are used.
As Rob said - making sure .dodeploy file is removed before we deploy anything no matter if
incremental and/or full is what is missing to cover all bases afaics.
when removing a module from AS runtime marker files are not removed
automatically
---------------------------------------------------------------------------------
Key: JBIDE-18862
URL:
https://issues.jboss.org/browse/JBIDE-18862
Project: Tools (JBoss Tools)
Issue Type: Feature Request
Components: server
Environment: Windiws7 64bit, JDK 8u25
Reporter: Darryl Miles
Priority: Minor
Fix For: 4.2.2.Final, 4.3.0.Alpha1
module removal does not delete the various marker files relating to it:
*.undeployed
*.dodeploy
The same should be true when ADDING a module, it should clean out any marker files.
Setup an deploy-inhibiting marker file to ensure it won't try to deploy it before the
*.ear is full written.
Write out the*.ear file.
Then adjust the marker files as necessary to reflect the intended state of the module (if
it should be started then remove the inhibiting marker file in the previous step).
--
This message was sent by Atlassian JIRA
(v6.3.8#6338)