[Geoserver-devel] CITE tests, legacy/web or web2?

Ben_Caradoc-Davies · March 27, 2009, 4:18am

Justin,

I am trying to run the CITE tests for GeoServer trunk, in preparation for CITE testing my GEOS-2753 WFS change. The legacy modules do not build (see below #2, some JAI and logging problems?).

- Should I run the CITE tests against legacy/web or community/web2?

- If I need it, what is the build procedure for web2?

- Is mvn jetty:run supported? (Running it at the top level results in an error. See below #1.)

- Has anyone run the CITE tests for GeoServer trunk in its current form?

- If I am targeting web2, is there support for building geoserver.war? Last time I asked there was not.

Updated developer documentation would be much appreciated!

Kind regards,
Ben.

****** Output of mvn jetty:run ******

[INFO] Scanning for projects...
[INFO] Reactor build order:
[INFO] GeoServer
[INFO] GeoServer Maven Plugins
[INFO] Configuration Deployment PlugIn
[INFO] GeoServer Maven Archetypes
[INFO] GeoServer WFS Output Format Archetype
[INFO] Core Platform Module
[INFO] Open Web Service Module
[INFO] Main Module
[INFO] Web Coverage Service Module
[INFO] Web Coverage Service 1.1 Module
[INFO] Web Feature Service Module
[INFO] Web Map Service Module
[INFO] GeoWebCache (GWC) Module
[INFO] Community Space
[INFO] GeoServer Extensions
[INFO] Searching repository for plugin with prefix: 'jetty'.
[INFO] ------------------------------------------------------------------------
[ERROR] BUILD ERROR
[INFO] ------------------------------------------------------------------------
[INFO] The plugin 'org.apache.maven.plugins:maven-jetty-plugin' does not exist o
r no valid version could be found
[INFO] ------------------------------------------------------------------------
[INFO] For more information, run Maven with the -e switch
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 2 seconds
[INFO] Finished at: Fri Mar 27 13:11:31 WST 2009
[INFO] Final Memory: 7M/14M
[INFO] ------------------------------------------------------------------------

****** Excerpt: mvn -Plegacy clean install ******

[INFO] Building Legacy Main Module
[INFO] task-segment: [clean, install]
[INFO] ------------------------------------------------------------------------
[INFO] [clean:clean]
[INFO] Deleting directory C:\Projects\car605\geoserver\cite\geoserver-trunk\src\
legacy\main\target
[INFO] [cobertura:clean {execution: default}]
[INFO] [resources:resources]
[INFO] Using default encoding to copy filtered resources.
[INFO] [compiler:compile]
[INFO] Compiling 53 source files to C:\Projects\car605\geoserver\cite\geoserver-
trunk\src\legacy\main\target\classes
[INFO] ------------------------------------------------------------------------
[ERROR] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Compilation failure

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[21,25] cannot find symbol
symbol : class JAIInfo
location: package org.geoserver.jai

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1023,4] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[57,11] org.vfny.geoserver.global.GeoSe
rver.ListenerWrapper is not abstract and does not override abstract method handl
ePostServiceChange(org.geoserver.config.ServiceInfo) in org.geoserver.config.Con
figurationListener

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[437,19] cannot find symbol
symbol : method getLoggingLevel()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[622,16] cannot find symbol
symbol : method setLoggingLevel(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[623,16] cannot find symbol
symbol : method setStdOutLogging(boolean)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[624,16] cannot find symbol
symbol : method setLoggingLocation(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[640,12] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[640,30] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[651,36] cannot find symbol
symbol : variable JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[994,19] cannot find symbol
symbol : method getLoggingLocation()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1003,12] cannot find symbol
symbol : method setLoggingLocation(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1011,20] cannot find symbol
symbol : method isStdOutLogging()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1019,12] cannot find symbol
symbol : method setStdOutLogging(boolean)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1024,16] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\
org\vfny\geoserver\global\GeoServer.java:[1024,49] cannot find symbol
symbol : variable JAIInfo
location: class org.vfny.geoserver.global.GeoServer

[INFO] ------------------------------------------------------------------------
[INFO] For more information, run Maven with the -e switch
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 41 seconds
[INFO] Finished at: Fri Mar 27 12:59:45 WST 2009
[INFO] Final Memory: 43M/82M
[INFO] ------------------------------------------------------------------------

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Justin_Deoliveira · March 27, 2009, 5:59am

Hi Ben,

Ben Caradoc-Davies wrote:

Justin,

I am trying to run the CITE tests for GeoServer trunk, in preparation for CITE testing my GEOS-2753 WFS change. The legacy modules do not build (see below #2, some JAI and logging problems?).

- Should I run the CITE tests against legacy/web or community/web2?

Either should work.

- If I need it, what is the build procedure for web2?

Just build with the "web2" profile

- Is mvn jetty:run supported? (Running it at the top level results in an error. See below #1.)

Not yet, we need to update web2/app/pom.xml and basically steal the configuration from legacy/web/pom.xml

- Has anyone run the CITE tests for GeoServer trunk in its current form?

We ran tests during the sprint after we ported all the services to the new configuration model and all tests passed, so hopefully things should not be too far off the mark. That said there has been some major hacking going on so you never know.

- If I am targeting web2, is there support for building geoserver.war? Last time I asked there was not.

Nope, but it is pretty easy. Basically we just have to change the packaging of web2/app/pom.xml to "war", rather than "jar". There may be a bit of other configuration to steal from legacy/web/pom.xml as well.

Updated developer documentation would be much appreciated!

I know... all the movement has trunk in a state that is undocumented. It is a pity. Things should hopefully stabilize soon and we can focus on updating the docs.

Kind regards,
Ben.

****** Output of mvn jetty:run ******

[INFO] Scanning for projects...
[INFO] Reactor build order:
[INFO] GeoServer
[INFO] GeoServer Maven Plugins
[INFO] Configuration Deployment PlugIn
[INFO] GeoServer Maven Archetypes
[INFO] GeoServer WFS Output Format Archetype
[INFO] Core Platform Module
[INFO] Open Web Service Module
[INFO] Main Module
[INFO] Web Coverage Service Module
[INFO] Web Coverage Service 1.1 Module
[INFO] Web Feature Service Module
[INFO] Web Map Service Module
[INFO] GeoWebCache (GWC) Module
[INFO] Community Space
[INFO] GeoServer Extensions
[INFO] Searching repository for plugin with prefix: 'jetty'.
[INFO] ------------------------------------------------------------------------
[ERROR] BUILD ERROR
[INFO] ------------------------------------------------------------------------
[INFO] The plugin 'org.apache.maven.plugins:maven-jetty-plugin' does not exist o
r no valid version could be found
[INFO] ------------------------------------------------------------------------
[INFO] For more information, run Maven with the -e switch
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 2 seconds
[INFO] Finished at: Fri Mar 27 13:11:31 WST 2009
[INFO] Final Memory: 7M/14M
[INFO] ------------------------------------------------------------------------

****** Excerpt: mvn -Plegacy clean install ******

[INFO] Building Legacy Main Module
[INFO] task-segment: [clean, install]
[INFO] ------------------------------------------------------------------------
[INFO] [clean:clean]
[INFO] Deleting directory C:\Projects\car605\geoserver\cite\geoserver-trunk\src\
legacy\main\target
[INFO] [cobertura:clean {execution: default}]
[INFO] [resources:resources]
[INFO] Using default encoding to copy filtered resources.
[INFO] [compiler:compile]
[INFO] Compiling 53 source files to C:\Projects\car605\geoserver\cite\geoserver-
trunk\src\legacy\main\target\classes
[INFO] ------------------------------------------------------------------------
[ERROR] BUILD FAILURE
[INFO] ------------------------------------------------------------------------
[INFO] Compilation failure

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[21,25] cannot find symbol
symbol : class JAIInfo
location: package org.geoserver.jai

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1023,4] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[57,11] org.vfny.geoserver.global.GeoSe
rver.ListenerWrapper is not abstract and does not override abstract method handl
ePostServiceChange(org.geoserver.config.ServiceInfo) in org.geoserver.config.Con
figurationListener

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[437,19] cannot find symbol
symbol : method getLoggingLevel()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[622,16] cannot find symbol
symbol : method setLoggingLevel(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[623,16] cannot find symbol
symbol : method setStdOutLogging(boolean)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[624,16] cannot find symbol
symbol : method setLoggingLocation(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[640,12] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[640,30] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[651,36] cannot find symbol
symbol : variable JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[994,19] cannot find symbol
symbol : method getLoggingLocation()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1003,12] cannot find symbol
symbol : method setLoggingLocation(java.lang.String)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1011,20] cannot find symbol
symbol : method isStdOutLogging()
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1019,12] cannot find symbol
symbol : method setStdOutLogging(boolean)
location: interface org.geoserver.config.GeoServerInfo

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1024,16] cannot find symbol
symbol : class JAIInfo
location: class org.vfny.geoserver.global.GeoServer

C:\Projects\car605\geoserver\cite\geoserver-trunk\src\legacy\main\src\main\java\

org\vfny\geoserver\global\GeoServer.java:[1024,49] cannot find symbol
symbol : variable JAIInfo
location: class org.vfny.geoserver.global.GeoServer

[INFO] ------------------------------------------------------------------------
[INFO] For more information, run Maven with the -e switch
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 41 seconds
[INFO] Finished at: Fri Mar 27 12:59:45 WST 2009
[INFO] Final Memory: 43M/82M
[INFO] ------------------------------------------------------------------------

--
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.

Ben_Caradoc-Davies · March 30, 2009, 3:17am

Justin,

thanks for your help. I will concentrate my efforts on web2, as it appears to be The Way Of The Future (TM). If I get jetty:run or war bundling to work, I will submit a patch.

How did you run the CITE tests if you have neither geoserver.war nor jetty:run support for web2? Manual deployment into WEB-INF?

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Andrea_Aime3 · March 30, 2009, 7:02am

Ben Caradoc-Davies ha scritto:

Justin,

thanks for your help. I will concentrate my efforts on web2, as it appears to be The Way Of The Future (TM). If I get jetty:run or war bundling to work, I will submit a patch.

How did you run the CITE tests if you have neither geoserver.war nor jetty:run support for web2? Manual deployment into WEB-INF?

Running GeoServer from within Eclipse using the Start application
that is provided in both the web module and in the web2 module,
among the test sources.
It starts an embedded Jetty with GeoServer.

It's the way I (most of us, I believe) usually develop against
GeoServer as well: one stop shop for quick start and live
debugging when you need it.

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Ben_Caradoc-Davies · March 30, 2009, 7:12am

Andrea Aime wrote:

Ben Caradoc-Davies ha scritto:

Justin,

thanks for your help. I will concentrate my efforts on web2, as it appears to be The Way Of The Future (TM). If I get jetty:run or war bundling to work, I will submit a patch.

How did you run the CITE tests if you have neither geoserver.war nor jetty:run support for web2? Manual deployment into WEB-INF?

Running GeoServer from within Eclipse using the Start application
that is provided in both the web module and in the web2 module,
among the test sources.
It starts an embedded Jetty with GeoServer.
It's the way I (most of us, I believe) usually develop against
GeoServer as well: one stop shop for quick start and live
debugging when you need it.

Excellent! This is the way I have always started GeoGerver web for debugging. I have never before had to mess with the Maven Jetty plugin (I was being led astray by the developers' manual). I should have been more specific in my request: I did not really want jetty:run, I wanted the Start application. If this is supported in web2, I am happy.

Thanks Andrea.

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Andrea_Aime3 · March 30, 2009, 7:17am

Ben Caradoc-Davies ha scritto:

Excellent! This is the way I have always started GeoGerver web for debugging. I have never before had to mess with the Maven Jetty plugin (I was being led astray by the developers' manual). I should have been more specific in my request: I did not really want jetty:run, I wanted the Start application. If this is supported in web2, I am happy.

Yup, it is. It is still good to have a full container way to start
GeoServer thought, in rare occasions a bug presents only stand-alone
due to the different way Eclipse and Maven setup the classpath
(Eclipse cannot tell a difference between a runtime and test dependency,
Maven does and feeds only the runtime ones to mvn:jetty).

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Ben_Caradoc-Davies · March 30, 2009, 7:34am

Andrea Aime wrote:

Ben Caradoc-Davies ha scritto:

Excellent! This is the way I have always started GeoGerver web for debugging. I have never before had to mess with the Maven Jetty plugin (I was being led astray by the developers' manual). I should have been more specific in my request: I did not really want jetty:run, I wanted the Start application. If this is supported in web2, I am happy.

Yup, it is. It is still good to have a full container way to start
GeoServer thought, in rare occasions a bug presents only stand-alone
due to the different way Eclipse and Maven setup the classpath
(Eclipse cannot tell a difference between a runtime and test dependency,
Maven does and feeds only the runtime ones to mvn:jetty).

Very true. I have had on occasion to overcome accidental dependency capture caused by the different dependency models of Eclipse and Maven. The principal incompatibility is, as you wrote, Eclipse having no distinction between runtime and test dependencies. One other problem is the handling of transitive dependencies; Eclipse includes workspace dependencies (not jar dependencies) on modules created in the same "mvn eclipse:eclipse" run, and so the capture of transitive dependencies also depends on how maven was run. In particular, trying to create a single Eclipse workspace with an aggregate pom to include both GeoServer and Geotools caused huge xml-api/xerces version problems. Separate maven runs act as kind of dependency firewall. I now use and recommend separate workspaces for GeoTools and GeoServer.

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Andrea_Aime3 · March 30, 2009, 7:37am

Ben Caradoc-Davies ha scritto:

Very true. I have had on occasion to overcome accidental dependency capture caused by the different dependency models of Eclipse and Maven. The principal incompatibility is, as you wrote, Eclipse having no distinction between runtime and test dependencies. One other problem is the handling of transitive dependencies; Eclipse includes workspace dependencies (not jar dependencies) on modules created in the same "mvn eclipse:eclipse" run, and so the capture of transitive dependencies also depends on how maven was run. In particular, trying to create a single Eclipse workspace with an aggregate pom to include both GeoServer and Geotools caused huge xml-api/xerces version problems. Separate maven runs act as kind of dependency firewall. I now use and recommend separate workspaces for GeoTools and GeoServer.

Yeah, been there me too, same conclusion.
Pity thought, I would like to be able and have GT2 and GS modules
depend on each other when I'm working on something that requires
changes on both (or changes in GT and checks in GS)

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Ben_Caradoc-Davies · March 30, 2009, 7:42am

Andrea Aime wrote:

Ben Caradoc-Davies ha scritto:

Very true. I have had on occasion to overcome accidental dependency capture caused by the different dependency models of Eclipse and Maven. The principal incompatibility is, as you wrote, Eclipse having no distinction between runtime and test dependencies. One other problem is the handling of transitive dependencies; Eclipse includes workspace dependencies (not jar dependencies) on modules created in the same "mvn eclipse:eclipse" run, and so the capture of transitive dependencies also depends on how maven was run. In particular, trying to create a single Eclipse workspace with an aggregate pom to include both GeoServer and Geotools caused huge xml-api/xerces version problems. Separate maven runs act as kind of dependency firewall. I now use and recommend separate workspaces for GeoTools and GeoServer.

Yeah, been there me too, same conclusion.
Pity thought, I would like to be able and have GT2 and GS modules
depend on each other when I'm working on something that requires
changes on both (or changes in GT and checks in GS)

Yes, it is a real pain not being able to use GeoTools hot code replacement when debugging GeoServer. The more ambitious I am with maven, the more I realise how little I know.

Regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Ben_Caradoc-Davies · March 30, 2009, 8:19am

Justin Deoliveira wrote:

Ben Caradoc-Davies wrote:

- If I am targeting web2, is there support for building geoserver.war? Last time I asked there was not.

Nope, but it is pretty easy. Basically we just have to change the packaging of web2/app/pom.xml to "war", rather than "jar". There may be a bit of other configuration to steal from legacy/web/pom.xml as well.

Justin, when I built GeoServer with -Pweb2, web-app/target contains a geoserver.war that looks suspiciously like the deployment bundle I will need. So perhaps it is fixed (perhaps not tested?).

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Andrea_Aime3 · March 30, 2009, 8:54am

Ben Caradoc-Davies ha scritto:

Justin,

thanks for your help. I will concentrate my efforts on web2, as it appears to be The Way Of The Future (TM). If I get jetty:run or war bundling to work, I will submit a patch.

How did you run the CITE tests if you have neither geoserver.war nor jetty:run support for web2? Manual deployment into WEB-INF?

Actually I just tried running jetty:run from web2/app and it does
work fine for me. I did something like:

mvn jetty:run -DGEOSERVER_DATA_DIR=..\..\..\..\data\release

from my community\web2\app module.

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Ben_Caradoc-Davies · March 30, 2009, 9:14am

Andrea Aime wrote:

Ben Caradoc-Davies ha scritto:

How did you run the CITE tests if you have neither geoserver.war nor jetty:run support for web2? Manual deployment into WEB-INF?

Actually I just tried running jetty:run from web2/app and it does
work fine for me. I did something like:
mvn jetty:run -DGEOSERVER_DATA_DIR=..\..\..\..\data\release
from my community\web2\app module.

Ha ha! Five minutes ago I finally discovered the correct number of ../ to include to get the wfs-1.0.0 configuration detected - by trial and error.

CITE wfs-1.0.0 tests are running now. I will report back.

Thanks Justin and Andrea for all your help.

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Ben_Caradoc-Davies · March 30, 2009, 9:31am

Ben Caradoc-Davies wrote:

CITE wfs-1.0.0 tests are running now. I will report back.

Most tests passed, but something failed. This is on a clean GeoTools r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753 patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

I found a huge number of target/logs/wfs-1.0.0 folders (one per test?). Is there an overall summary report that might help me find the test(s) that failed? Or is "grep -H 'result=' */*" as good as it gets?

What is the recommended heap size for running the test engine?

Kind regards,

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Andrea_Aime3 · March 30, 2009, 10:13am

Ben Caradoc-Davies ha scritto:

Ben Caradoc-Davies wrote:

CITE wfs-1.0.0 tests are running now. I will report back.

Most tests passed, but something failed. This is on a clean GeoTools r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753 patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

I found a huge number of target/logs/wfs-1.0.0 folders (one per test?). Is there an overall summary report that might help me find the test(s) that failed? Or is "grep -H 'result=' */*" as good as it gets?

If you look in README.txt there are a few command line suggestions
on how to drill down the logs.
As far as I'm concerned I usually modify the pom to only
include the scripts I need, build, look for the generated war
(it's somewhere in the cite engine directory) and deploy
the war into tomcat.
That allows me to run the tests multiple times, rerun just a single
test, and drill down the issues with the web UI.
But make sure you cherry pick the script you need, otherwise
the container will OOM when that stupid beast tries to xslt
process all the scripts.
Also, you cannot rename the generated war. So what I do, and
be seated before reading, is to keep a separate tomcat instance
for each of the compiled cite test suites (one war for each)
(I also modified them to run on 9090).
Gross, but works.

Cheers
Andrea

--
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

Justin_Deoliveira · March 30, 2009, 3:28pm

Ben Caradoc-Davies wrote:

Justin Deoliveira wrote:

Ben Caradoc-Davies wrote:

- If I am targeting web2, is there support for building geoserver.war? Last time I asked there was not.

Nope, but it is pretty easy. Basically we just have to change the packaging of web2/app/pom.xml to "war", rather than "jar". There may be a bit of other configuration to steal from legacy/web/pom.xml as well.

Justin, when I built GeoServer with -Pweb2, web-app/target contains a geoserver.war that looks suspiciously like the deployment bundle I will need. So perhaps it is fixed (perhaps not tested?).

Yup, looks like I was wrong. When I said it did not I looked at web2/pom.xml but of course everything "webapp" related is not in web2/app/pom.xml. Apoliogies for the misdirection.

Kind regards,

--
Justin Deoliveira
OpenGeo - http://opengeo.org
Enterprise support for open source geospatial.

Ben_Caradoc-Davies · March 31, 2009, 4:15am

Ben Caradoc-Davies wrote:

Most tests passed, but something failed. This is on a clean GeoTools r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753 patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

Justin,

I have 12 failures running CITE wfs-1.0.0 on unmodified geoserver trunk. I have attached the log for the first of these. I do not know what is wrong. I built the test engine with Java 5, but am running it with 6.

Can anyone confirm these failures? (This is my first CITE run.)

Kind regards,
Ben.

>log.bat wfs 1.0.0 | perl -ne "print if /Failed/"
Test wfs:main (wfs-1.0.0) Failed (Inherited Failure)
    Test wfs:test1.0.0-basic-describefeaturetype-get-1 (wfs-1.0.0/w28aac25b3b9d169b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-get-2 (wfs-1.0.0/w28aac25b3b9d171b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-get-4 (wfs-1.0.0/w28aac25b3b9d175b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-get-5 (wfs-1.0.0/w28aac25b3b9d177b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-get-6 (wfs-1.0.0/w28aac25b3b9d179b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-get-7 (wfs-1.0.0/w28aac25b3b9d181b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-1 (wfs-1.0.0/w28aac25b3b9d187b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-2 (wfs-1.0.0/w28aac25b3b9d189b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-4 (wfs-1.0.0/w28aac25b3b9d193b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-5 (wfs-1.0.0/w28aac25b3b9d195b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-6 (wfs-1.0.0/w28aac25b3b9d197b1_1) Failed
    Test wfs:test1.0.0-basic-describefeaturetype-post-7 (wfs-1.0.0/w28aac25b3b9d199b1_1) Failed

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

(attachments)

log.xml (3.88 KB)

jive · March 31, 2009, 4:44am

Ben the log you attached indicates:
<param name="VAR_WFS_DESCRIBE_FEATURE_TYPE_HTTP_GET_URL" label="<>">
<value>http://localhost:8080/geoserver/wfs?request=DescribeFeatureType</value>
</param>

As such the cite server would be checking itself; can you fix up your
configuration and try again?
Jody

On Tue, Mar 31, 2009 at 3:15 PM, Ben Caradoc-Davies
<Ben.Caradoc-Davies@anonymised.com> wrote:

Ben Caradoc-Davies wrote:

Most tests passed, but something failed. This is on a clean GeoTools
r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753
patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

Justin,

I have 12 failures running CITE wfs-1.0.0 on unmodified geoserver trunk. I
have attached the log for the first of these. I do not know what is wrong. I
built the test engine with Java 5, but am running it with 6.

Can anyone confirm these failures? (This is my first CITE run.)

Kind regards,
Ben.

log.bat wfs 1.0.0 | perl -ne "print if /Failed/"

Test wfs:main (wfs-1.0.0) Failed (Inherited Failure)
Test wfs:test1.0.0-basic-describefeaturetype-get-1
(wfs-1.0.0/w28aac25b3b9d169b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-2
(wfs-1.0.0/w28aac25b3b9d171b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-4
(wfs-1.0.0/w28aac25b3b9d175b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-5
(wfs-1.0.0/w28aac25b3b9d177b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-6
(wfs-1.0.0/w28aac25b3b9d179b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-7
(wfs-1.0.0/w28aac25b3b9d181b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-1
(wfs-1.0.0/w28aac25b3b9d187b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-2
(wfs-1.0.0/w28aac25b3b9d189b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-4
(wfs-1.0.0/w28aac25b3b9d193b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-5
(wfs-1.0.0/w28aac25b3b9d195b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-6
(wfs-1.0.0/w28aac25b3b9d197b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-7
(wfs-1.0.0/w28aac25b3b9d199b1_1) Failed

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

------------------------------------------------------------------------------

_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

Ben_Caradoc-Davies · March 31, 2009, 5:07am

Sorry, Jody, I do not understand. Please clarify. How is the server testing itself? I thought the cite engine was testing the server (GeoServer) by issuing WFS requests and analysing responses. All other tests pass, and I can see sensible GeoServer console output in Eclipse.

I am running the CITE wfs 1.0.0 test, as documented here:
http://gridlock.openplans.org/geoserver/1.7.x/doc/developer/cite-test-guide/index.html

As instructed, I have in the cite test GUI set the location of the capabilities document as this:
http://localhost:8080/geoserver/wfs?service=wfs&request=getcapabilities&version=1.0.0
(I am testing a local GeoServer responding on this port.)

My first observation is that the log.xml has no FeatureType. The returned schema has no element tags. Is this correct?

Kind regards,
Ben.

Jody Garnett wrote:

Ben the log you attached indicates:
<param name="VAR_WFS_DESCRIBE_FEATURE_TYPE_HTTP_GET_URL" label="<>">
  <value>http://localhost:8080/geoserver/wfs?request=DescribeFeatureType</value>
  </param>

As such the cite server would be checking itself; can you fix up your
configuration and try again?
Jody

On Tue, Mar 31, 2009 at 3:15 PM, Ben Caradoc-Davies
<Ben.Caradoc-Davies@anonymised.com> wrote:

Ben Caradoc-Davies wrote:

Most tests passed, but something failed. This is on a clean GeoTools
r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753
patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

Justin,

I have 12 failures running CITE wfs-1.0.0 on unmodified geoserver trunk. I
have attached the log for the first of these. I do not know what is wrong. I
built the test engine with Java 5, but am running it with 6.

Can anyone confirm these failures? (This is my first CITE run.)

Kind regards,
Ben.

log.bat wfs 1.0.0 | perl -ne "print if /Failed/"

Test wfs:main (wfs-1.0.0) Failed (Inherited Failure)
  Test wfs:test1.0.0-basic-describefeaturetype-get-1
(wfs-1.0.0/w28aac25b3b9d169b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-get-2
(wfs-1.0.0/w28aac25b3b9d171b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-get-4
(wfs-1.0.0/w28aac25b3b9d175b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-get-5
(wfs-1.0.0/w28aac25b3b9d177b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-get-6
(wfs-1.0.0/w28aac25b3b9d179b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-get-7
(wfs-1.0.0/w28aac25b3b9d181b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-1
(wfs-1.0.0/w28aac25b3b9d187b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-2
(wfs-1.0.0/w28aac25b3b9d189b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-4
(wfs-1.0.0/w28aac25b3b9d193b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-5
(wfs-1.0.0/w28aac25b3b9d195b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-6
(wfs-1.0.0/w28aac25b3b9d197b1_1) Failed
  Test wfs:test1.0.0-basic-describefeaturetype-post-7
(wfs-1.0.0/w28aac25b3b9d199b1_1) Failed

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

------------------------------------------------------------------------------

_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

jive · March 31, 2009, 5:45am

Have a look at the log.xml your provided; now I could be reading it
wrong; but it looks like the param value for
VAR_WFS_DESCRIBE_FEATURE_TYPE_HTTP_GET_URL is wrong. Have a look and
let me know what you think.

I think this should point to your server from the perspective of the
CITE test engine. Are you running CITE locally? I see the file =
"C:\Projects\car605\geoserver\cite\cite\tests\wfs\1.0.0\ets\ctl\wfs.xml"
mentioned so perhaps you are running it locally.

Jody

On Tue, Mar 31, 2009 at 4:07 PM, Ben Caradoc-Davies
<Ben.Caradoc-Davies@anonymised.com> wrote:

Sorry, Jody, I do not understand. Please clarify. How is the server testing
itself? I thought the cite engine was testing the server (GeoServer) by
issuing WFS requests and analysing responses. All other tests pass, and I
can see sensible GeoServer console output in Eclipse.

I am running the CITE wfs 1.0.0 test, as documented here:
http://gridlock.openplans.org/geoserver/1.7.x/doc/developer/cite-test-guide/index.html

As instructed, I have in the cite test GUI set the location of the
capabilities document as this:
http://localhost:8080/geoserver/wfs?service=wfs&request=getcapabilities&version=1.0.0
(I am testing a local GeoServer responding on this port.)

My first observation is that the log.xml has no FeatureType. The returned
schema has no element tags. Is this correct?

Kind regards,
Ben.

Jody Garnett wrote:

Ben the log you attached indicates:
<param name="VAR_WFS_DESCRIBE_FEATURE_TYPE_HTTP_GET_URL" label="<>">

<value>http://localhost:8080/geoserver/wfs?request=DescribeFeatureType</value>
</param>

As such the cite server would be checking itself; can you fix up your
configuration and try again?
Jody

On Tue, Mar 31, 2009 at 3:15 PM, Ben Caradoc-Davies
<Ben.Caradoc-Davies@anonymised.com> wrote:

Ben Caradoc-Davies wrote:

Most tests passed, but something failed. This is on a clean GeoTools
r32719 / GeoServer r11894 checkout, testing web2, *before* my GEOS-2753
patch is applied. The failure was:
Test wfs:main Failed (Inherited Failure)

Justin,

I have 12 failures running CITE wfs-1.0.0 on unmodified geoserver trunk.
I
have attached the log for the first of these. I do not know what is
wrong. I
built the test engine with Java 5, but am running it with 6.

Can anyone confirm these failures? (This is my first CITE run.)

Kind regards,
Ben.

log.bat wfs 1.0.0 | perl -ne "print if /Failed/"

Test wfs:main (wfs-1.0.0) Failed (Inherited Failure)
Test wfs:test1.0.0-basic-describefeaturetype-get-1
(wfs-1.0.0/w28aac25b3b9d169b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-2
(wfs-1.0.0/w28aac25b3b9d171b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-4
(wfs-1.0.0/w28aac25b3b9d175b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-5
(wfs-1.0.0/w28aac25b3b9d177b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-6
(wfs-1.0.0/w28aac25b3b9d179b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-get-7
(wfs-1.0.0/w28aac25b3b9d181b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-1
(wfs-1.0.0/w28aac25b3b9d187b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-2
(wfs-1.0.0/w28aac25b3b9d189b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-4
(wfs-1.0.0/w28aac25b3b9d193b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-5
(wfs-1.0.0/w28aac25b3b9d195b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-6
(wfs-1.0.0/w28aac25b3b9d197b1_1) Failed
Test wfs:test1.0.0-basic-describefeaturetype-post-7
(wfs-1.0.0/w28aac25b3b9d199b1_1) Failed

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

------------------------------------------------------------------------------

_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia

Ben_Caradoc-Davies · March 31, 2009, 6:04am

Jody Garnett wrote:

Have a look at the log.xml your provided; now I could be reading it
wrong; but it looks like the param value for
VAR_WFS_DESCRIBE_FEATURE_TYPE_HTTP_GET_URL is wrong. Have a look and
let me know what you think.

The URL is http://localhost:8080/geoserver/wfs?request=DescribeFeatureType
This looks OK to me, except that it has no typename, nor is one present in the list of parameters in the the cite log request section.

If I submit a URL with a typename:
http://localhost:8080/geoserver/wfs?service=WFS&request=DescribeFeatureType&version=1.0.0&typeName=cdf:Fifteen
I get an XML Schema response that includes

<xs:complexType name="Fifteen_Type">
...

and

<xs:element name="Fifteen" type="cdf:FifteenType" substitutionGroup="gml:_Feature"/>

If I omit the typename parameter and submit the following URL in my web browser:
http://localhost:8080/geoserver/wfs?service=WFS&request=DescribeFeatureType&version=1.0.0
I get a response that is an XML Schema that has only imports and no other content. I think this is what the cite tests are complaining about.

What is the expected behaviour of a WFS in response to a DescribeFeatureType with no TypeName?

I think this should point to your server from the perspective of the
CITE test engine. Are you running CITE locally? I see the file =
"C:\Projects\car605\geoserver\cite\cite\tests\wfs\1.0.0\ets\ctl\wfs.xml"
mentioned so perhaps you are running it locally.

Yes, I am running GeoServer locally.

--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineer, CSIRO Exploration and Mining
Australian Resources Research Centre
26 Dick Perry Ave, Kensington WA 6151, Australia