This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
With the current Q build (netbeans-5_5-qbuild-bin-200608181330-windows-18_Aug_2006_1330.exe) I am having a problem not seen in NB 5.5 beta 1. If I use the Sun Java System Application Server's own commands to start the AppServer from outside NetBeans, then when I try to deploy an application in NetBeans, I get an error: Another process is using the administrative port: 4,848 NB is trying to stop and restart the AppServer before deploying the app. It used to perform this task, and it always succeeded in doing so. Now it gets this error and fails. Now I can only deploy an app to the AppServer if I don't start the AppServer outside of NB. NB only recognizes the AppServer if NB started it. This is a significant regression from previous builds. The Tutorial has always provided instructions on how to start and stop the AppServer. If this problem is not fixed, we will have to explicitly state that users MUST NOT start the AppServer if they will be using NetBeans, only if they will be using Ant. This seems to expose a bug in NetBeans, which seems undesirable.
regression from a fix related to profiler bug. Vince, take a look.
I just did a clean check-out and build of release55 on WinXP and ran it against jdk1.5.0_06. I am using GF v1 u1 b5, which I started outside the IDE. I created an ent app that has a simple web app in it. I used Debug Main Project to force a stop/start and could not replicate the issue described by the filer. Need more info to duplicate. Server build? JDK in use by NB and server? output window content for the deploy command? relevant server log content? Project that demonstrates this issue? Please reopen with more detailed information, if this issue persists in recent daily builds...
Cannot reproduce using latest daily build (netbeans-5_5-daily-bin-200609050000-windows-5_Sep_2006_0000.exe). Was using server build sjsas_pe-9_0-fcs-bin-b48-windows-04_may_2006.exe with JDK 1.5.0_06. Project that demonstrated issue was the Java EE tutorial example examples/web/hello1 (the first application in the Tutorial, in Chapter 2).
Vince: the user was able to see this behaviour. The check was added recently, and I am not sure why. We have extensive uimpl for isRunning() which is much more than a socket lookup. Not reproducible yet does not mean we can ignore this bug. Reopening, to see how we can avoid this check to that it will never happen...
portInUse check was added to fail fast, if the server would not be able to start due to some other process running on port the admin port. The most common case is when the user has multiple servers registered on the same admin port.... If you have a registered instance A with admin port Z running and then try to start registered instance B, that also uses port Z, the start operation for registered instance B takes a long time to fail and the message about the failure is full of false positives for the root cause... This change is actually related to issue 81422
it looks like ludo fixed this, by commenting out the message