I’ve been having test failures locally on geoserver trunk for some time, but since I am not all that active there and no one else seems to be complaining I didn’t bring it up on the list. Today I finally found some time to investigate, and actually found two issues. Now I get all tests passing
xerces version conflict - For most modules in GeoServer is identifying xercesImpl 2.6.2 as the correct one to include on the classpath, however for org.geoserver:main it is using 2.4.0. This is causing AbstractMethodError’s at runtime, as you might expect. I tried bumping the version in geoserver’s main pom to 2.6.2 and tests pass, any objections to committing it?
networking in coverage setup - this one is a little weird… redhat based systems like my fedora laptop have a weird glitch where setting a custom hostname during the install process doesn’t necessarily update the /etc/hosts file, then some programs have issues when “myname.localdomain” doesn’t resolve. The JVM seems to be one of those programs. Ok, so I fixed my hosts file. The weird part is that I was running into those errors during setup of coveragestores in mock data directories for GeoServer’s unit tests, should there be networking going on during this operation? I’m attaching a stack trace for one of the failing tests to this mail.
Thanks David; I was just stressing over the xerces version conflict (trying to test geotools patches against geoserver). Please commit
It is hard to cut down on networking going on during units tests; some classes like URL like to hit the network for hashcode and equals etc…
–
Jody Garnett
On Sunday, 19 June 2011 at 12:50 PM, David Winslow wrote:
Hi all,
I’ve been having test failures locally on geoserver trunk for some time, but since I am not all that active there and no one else seems to be complaining I didn’t bring it up on the list. Today I finally found some time to investigate, and actually found two issues. Now I get all tests passing
xerces version conflict - For most modules in GeoServer is identifying xercesImpl 2.6.2 as the correct one to include on the classpath, however for org.geoserver:main it is using 2.4.0. This is causing AbstractMethodError’s at runtime, as you might expect. I tried bumping the version in geoserver’s main pom to 2.6.2 and tests pass, any objections to committing it?
networking in coverage setup - this one is a little weird… redhat based systems like my fedora laptop have a weird glitch where setting a custom hostname during the install process doesn’t necessarily update the /etc/hosts file, then some programs have issues when “myname.localdomain” doesn’t resolve. The JVM seems to be one of those programs. Ok, so I fixed my hosts file. The weird part is that I was running into those errors during setup of coveragestores in mock data directories for GeoServer’s unit tests, should there be networking going on during this operation? I’m attaching a stack trace for one of the failing tests to this mail.
EditLive Enterprise is the world’s most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev
Looking into the xerces version hell for a bit; the root pom explicitly has 2.4.0 in the dependency management section.
GeoTools meanwhile has 2.7.1 - so my guess is most of the modules get xerces as a transitive dependency; and main that actually uses it directly is the odd fish out.
So this appears to be a long standing issue? Any idea why it is hitting us today…
–
Jody Garnett
On Sunday, 19 June 2011 at 3:06 PM, Jody Garnett wrote:
Thanks David; I was just stressing over the xerces version conflict (trying to test geotools patches against geoserver). Please commit
It is hard to cut down on networking going on during units tests; some classes like URL like to hit the network for hashcode and equals etc…
–
Jody Garnett
On Sunday, 19 June 2011 at 12:50 PM, David Winslow wrote:
Hi all,
I’ve been having test failures locally on geoserver trunk for some time, but since I am not all that active there and no one else seems to be complaining I didn’t bring it up on the list. Today I finally found some time to investigate, and actually found two issues. Now I get all tests passing
xerces version conflict - For most modules in GeoServer is identifying xercesImpl 2.6.2 as the correct one to include on the classpath, however for org.geoserver:main it is using 2.4.0. This is causing AbstractMethodError’s at runtime, as you might expect. I tried bumping the version in geoserver’s main pom to 2.6.2 and tests pass, any objections to committing it?
networking in coverage setup - this one is a little weird… redhat based systems like my fedora laptop have a weird glitch where setting a custom hostname during the install process doesn’t necessarily update the /etc/hosts file, then some programs have issues when “myname.localdomain” doesn’t resolve. The JVM seems to be one of those programs. Ok, so I fixed my hosts file. The weird part is that I was running into those errors during setup of coveragestores in mock data directories for GeoServer’s unit tests, should there be networking going on during this operation? I’m attaching a stack trace for one of the failing tests to this mail.
EditLive Enterprise is the world’s most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have one failure in the flow control extension:
Test set: org.geoserver.flow.controller.UserFlowControllerTest
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec <<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest) Time elapsed: 1.212 sec <<< FAILURE!
junit.framework.AssertionFailedError: expected: but was:
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
–
Jody Garnett
On Sunday, 19 June 2011 at 3:26 PM, Jody Garnett wrote:
Looking into the xerces version hell for a bit; the root pom explicitly has 2.4.0 in the dependency management section.
GeoTools meanwhile has 2.7.1 - so my guess is most of the modules get xerces as a transitive dependency; and main that actually uses it directly is the odd fish out.
So this appears to be a long standing issue? Any idea why it is hitting us today…
–
Jody Garnett
On Sunday, 19 June 2011 at 3:06 PM, Jody Garnett wrote:
Thanks David; I was just stressing over the xerces version conflict (trying to test geotools patches against geoserver). Please commit
It is hard to cut down on networking going on during units tests; some classes like URL like to hit the network for hashcode and equals etc…
–
Jody Garnett
On Sunday, 19 June 2011 at 12:50 PM, David Winslow wrote:
Hi all,
I’ve been having test failures locally on geoserver trunk for some time, but since I am not all that active there and no one else seems to be complaining I didn’t bring it up on the list. Today I finally found some time to investigate, and actually found two issues. Now I get all tests passing
xerces version conflict - For most modules in GeoServer is identifying xercesImpl 2.6.2 as the correct one to include on the classpath, however for org.geoserver:main it is using 2.4.0. This is causing AbstractMethodError’s at runtime, as you might expect. I tried bumping the version in geoserver’s main pom to 2.6.2 and tests pass, any objections to committing it?
networking in coverage setup - this one is a little weird… redhat based systems like my fedora laptop have a weird glitch where setting a custom hostname during the install process doesn’t necessarily update the /etc/hosts file, then some programs have issues when “myname.localdomain” doesn’t resolve. The JVM seems to be one of those programs. Ok, so I fixed my hosts file. The weird part is that I was running into those errors during setup of coveragestores in mock data directories for GeoServer’s unit tests, should there be networking going on during this operation? I’m attaching a stack trace for one of the failing tests to this mail.
EditLive Enterprise is the world’s most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking. http://p.sf.net/sfu/ephox-dev2dev
On Sun, Jun 19, 2011 at 8:05 AM, Jody Garnett <jody.garnett@anonymised.com> wrote:
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have
one failure in the flow control extension:
-------------------------------------------------------------------------------
Test set: org.geoserver.flow.controller.UserFlowControllerTest
-------------------------------------------------------------------------------
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec
<<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest)
Time elapsed: 1.212 sec <<< FAILURE!
junit.framework.AssertionFailedError: expected:<PROCESSING> but
was:<STARTED>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at
org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
Yeah, these random exception occur every now and then on other
systems... but not on my machines and very rarely on the build box...
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
Yes, I saw this behavior on oracle jvm 1.6.0_26 and whichever openjdk
release is current in fedora (I'm not currently at my computer.) I
haven't tried with a java 5 setup since I started having trouble on
this machine.
On Sunday, June 19, 2011, Andrea Aime <andrea.aime@anonymised.com> wrote:
On Sun, Jun 19, 2011 at 8:05 AM, Jody Garnett <jody.garnett@anonymised.com> wrote:
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have
one failure in the flow control extension:
-------------------------------------------------------------------------------
Test set: org.geoserver.flow.controller.UserFlowControllerTest
-------------------------------------------------------------------------------
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec
<<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest)
Time elapsed: 1.212 sec <<< FAILURE!
junit.framework.AssertionFailedError: expected:<PROCESSING> but
was:<STARTED>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at
org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
Yeah, these random exception occur every now and then on other
systems... but not on my machines and very rarely on the build box...
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
Yes, I saw this behavior on oracle jvm 1.6.0_26 and whichever openjdk
release is current in fedora (I'm not currently at my computer.) I
haven't tried with a java 5 setup since I started having trouble on
this machine.
On Sunday, June 19, 2011, Andrea Aime <andrea.aime@anonymised.com> wrote:
On Sun, Jun 19, 2011 at 8:05 AM, Jody Garnett <jody.garnett@anonymised.com> wrote:
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have
one failure in the flow control extension:
-------------------------------------------------------------------------------
Test set: org.geoserver.flow.controller.UserFlowControllerTest
-------------------------------------------------------------------------------
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec
<<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest)
Time elapsed: 1.212 sec <<< FAILURE!
junit.framework.AssertionFailedError: expected:<PROCESSING> but
was:<STARTED>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at
org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
Yeah, these random exception occur every now and then on other
systems... but not on my machines and very rarely on the build box...
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
Yes, I saw this behavior on oracle jvm 1.6.0_26 and whichever openjdk
release is current in fedora (I'm not currently at my computer.) I
haven't tried with a java 5 setup since I started having trouble on
this machine.
On Sunday, June 19, 2011, Andrea Aime <andrea.aime@anonymised.com> wrote:
On Sun, Jun 19, 2011 at 8:05 AM, Jody Garnett <jody.garnett@anonymised.com> wrote:
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have
one failure in the flow control extension:
-------------------------------------------------------------------------------
Test set: org.geoserver.flow.controller.UserFlowControllerTest
-------------------------------------------------------------------------------
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec
<<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest)
Time elapsed: 1.212 sec <<< FAILURE!
junit.framework.AssertionFailedError: expected:<PROCESSING> but
was:<STARTED>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at
org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
Yeah, these random exception occur every now and then on other
systems... but not on my machines and very rarely on the build box...
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
Jody, here is the bug report I filed for this issue:
GEOS-4462: Intermittent control-flow {User,Global}FlowControllerTest failures break build http://jira.codehaus.org/browse/GEOS-4462
On 19/06/11 14:57, Andrea Aime wrote:
On Sun, Jun 19, 2011 at 8:05 AM, Jody Garnett<jody.garnett@anonymised.com> wrote:
After you hint to use xerces 2.7.1 most of geoserver builds for me; I have
one failure in the flow control extension:
-------------------------------------------------------------------------------
Test set: org.geoserver.flow.controller.UserFlowControllerTest
-------------------------------------------------------------------------------
Tests run: 1, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 1.226 sec
<<< FAILURE!
testConcurrentRequestsSingleUser(org.geoserver.flow.controller.UserFlowControllerTest)
Time elapsed: 1.212 sec<<< FAILURE!
junit.framework.AssertionFailedError: expected:<PROCESSING> but
was:<STARTED>
at junit.framework.Assert.fail(Assert.java:47)
at junit.framework.Assert.failNotEquals(Assert.java:280)
at junit.framework.Assert.assertEquals(Assert.java:64)
at junit.framework.Assert.assertEquals(Assert.java:71)
at
org.geoserver.flow.controller.UserFlowControllerTest.testConcurrentRequestsSingleUser(UserFlowControllerTest.java:49)
Yeah, these random exception occur every now and then on other
systems... but not on my machines and very rarely on the build box...
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
Ben Caradoc-Davies <Ben.Caradoc-Davies@anonymised.com>
Software Engineering Team Leader
CSIRO Earth Science and Resource Engineering
Australian Resources Research Centre
As for Xerces, I believe you are both building with java 6 already?
I found that removing the dependency solid (from the main pom) made
the build work fine, too
Cheers
Andrea
--
-------------------------------------------------------
Ing. Andrea Aime
GeoSolutions S.A.S.
Tech lead
Via Poggio alle Viti 1187
55054 Massarosa (LU)
Italy
