Comment 1 Angel Vera 2010-10-28 15:46:10 EDT

I ran the test locally a few times with the 3.2.3 build and I didn't get any failures. Looking at the out logs, I see the following errors which seems to be related to the changes I did recently for Bug# 328067:

ENTRY org.eclipse.core.jobs 4 2 2010-10-28 16:28:09.970
!MESSAGE An internal error occurred during: "Starting Tomcat v5.0 Server at localhost".
!STACK 0
java.lang.NoClassDefFoundError: org/eclipse/debug/ui/DebugUITools
		 at org.eclipse.wst.server.ui.internal.SaveEditorPrompter.setDebugSaveBeforeLaunching(SaveEditorPrompter.java:62)
		 at org.eclipse.wst.server.ui.internal.SaveEditorPrompter.setDebugNeverSave(SaveEditorPrompter.java:69)
		 at org.eclipse.wst.server.core.internal.Server.startImpl2(Server.java:3213)
		 at org.eclipse.wst.server.core.internal.Server.startImpl(Server.java:3163)
		 at org.eclipse.wst.server.core.internal.Server$StartJob.run(Server.java:359)
		 at org.eclipse.core.internal.jobs.Worker.run(Worker.java:54)

Comment 2 David Williams 2010-10-28 17:38:02 EDT

I don't quite recall the history (had to do with not shipping "java source" in binary folders, I believe) but jst.server.tomcat.core has some "pre-compiled jars" in it, I believe. See 

org.eclipse.jst.server.tomcat.runtime.50.loader-src.zip
org.eclipse.jst.server.tomcat.runtime.50.loader.jar
org.eclipse.jst.server.tomcat.runtime.55.loader-src.zip
org.eclipse.jst.server.tomcat.runtime.55.loader.jar
org.eclipse.jst.server.tomcat.runtime.60.loader-src.zip
org.eclipse.jst.server.tomcat.runtime.60.loader.jar
org.eclipse.jst.server.tomcat.runtime.70.loader-src.zip
org.eclipse.jst.server.tomcat.runtime.70.loader.jar

Do these need to be recompiled with the recent changes, checked into cvs, and released?

Comment 3 Larry Isaacs 2010-10-28 22:57:35 EDT

FYI: The org.eclipse.jst.server.tomcat.runtime.??.loader jars are copied to Tomcat as part of the "Serve modules without publishing" feature.  Their code only runs within Tomcat and wouldn't affect anything or be affected by anything running in Eclipse.

Comment 4 Angel Vera 2010-11-02 11:39:32 EDT

Created attachment 182219 [details]
logs from the junit run

Darin, 

Can you shed some light on this jUnit failure. This is related to the changes that you and I discussed for bug# 328067. 

The code changes are in Server.java ver. 1.188 as per: http://dev.eclipse.org/viewcvs/index.cgi/servertools/plugins/org.eclipse.wst.server.core/servercore/org/eclipse/wst/server/core/internal/Server.java?root=WebTools_Project&r1=1.187&r2=1.188

The jUnit is defined as a <ant target="core-test", which I think should be the right target, but why is my code detecting that org.eclipse.ui is active, when it shouldn't be.

Comment 5 Angel Vera 2010-11-04 13:10:41 EDT

Darin, I missed adding you on my last comment. Do you have any ideas, this is related to the hang that we worked on bug# 328067. Take a look at comment #4

Comment 6 Angel Vera 2010-11-04 14:58:39 EDT

*** Bug 329475 has been marked as a duplicate of this bug. ***

Comment 7 Darin Wright 2010-11-04 17:19:59 EDT

(In reply to comment #4)
> The jUnit is defined as a <ant target="core-test", which I think should be the
> right target, but why is my code detecting that org.eclipse.ui is active, when
> it shouldn't be.

I'm not familiar with the code that checks if the UI is running. Can you debug through the test to see why it thinks the UI is running?

Comment 8 Angel Vera 2010-11-08 11:38:11 EST

(In reply to comment #7)
Unfortunately, this is happening in our automated jUnit test and I don't know how I could debug that. Maybe if I setup the same environment but that might take sometime.

Comment 9 Angel Vera 2010-11-08 13:07:58 EST

I committed changes to 32M to make the test UI, I am doing this as a workaround to the jUnit test. While we continue to investigate the problem

Comment 10 Angel Vera 2010-11-08 13:10:11 EST

John, 

Do you know of any better of way of checking if the UI is enabled or not?

Comment 11 Angel Vera 2010-11-11 15:48:26 EST

Sicne the changes for 32M work as a temp workaround. I am committed and released the same changes for HEAD(3.3) while we investigate where the failure comes from.

Comment 12 Troy Bishop 2010-11-30 16:35:30 EST

Created attachment 184189 [details]
possible patch

Within a WTP adopter product we have support for headlessly starting a server and this problem is now occurring.  Attached is a patch which modifies the check for determining if in a headless or UI environment.  It solves the problem in my environment.

Comment 13 Angel Vera 2010-12-07 11:30:06 EST

Thanks for the patch Troy, and the testing. It seems to do the trick. It still puzzles me why the the .ui plugin is detected as active but that will be a problem for a different time.

Comment 14 Angel Vera 2010-12-07 11:30:18 EST

committed to 32M

Comment 15 Angel Vera 2010-12-08 16:08:04 EST

committed to HEAD

Comment 16 Angel Vera 2010-12-08 16:08:36 EST

release to HEAD

Comment 17 Angel Vera 2010-12-08 16:08:52 EST

released to 32M

Comment 18 Eclipse Genie 2017-10-11 16:35:02 EDT

New Gerrit change created: https://git.eclipse.org/r/109002

Comment 19 Eclipse Genie 2017-10-11 16:35:05 EDT

New Gerrit change created: https://git.eclipse.org/r/108999

Comment 20 Eclipse Genie 2017-10-11 16:35:13 EDT

New Gerrit change created: https://git.eclipse.org/r/109003

Comment 21 Eclipse Genie 2017-10-11 16:54:01 EDT

New Gerrit change created: https://git.eclipse.org/r/109836

Comment 22 Eclipse Genie 2017-10-11 16:54:03 EDT

New Gerrit change created: https://git.eclipse.org/r/109835

Comment 23 Eclipse Genie 2017-10-11 16:57:02 EDT

Gerrit change https://git.eclipse.org/r/109836 was merged to [master].
Commit: http://git.eclipse.org/c/servertools/webtools.servertools.git/commit/?id=83499cd59a24d79fc4f2fdd2765af74c00b3dbe6

Comment 24 Eclipse Genie 2017-10-11 16:57:06 EDT

Gerrit change https://git.eclipse.org/r/109835 was merged to [master].
Commit: http://git.eclipse.org/c/servertools/webtools.servertools.git/commit/?id=a0e753e6dc23ae8123f1b15256a922d4d9508472