openjpa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Patrick Linskey" <plins...@gmail.com>
Subject Re: Failures in TeamCity Derby build and svn commit: r616972
Date Fri, 01 Feb 2008 01:52:58 GMT
> If the hangup is intolerable, let me know, and I'll back out my change
> tomorrow.

Personally, I'd rather have the TeamCity build be clean while you
investigate. As you noted, the error-reporting from surefire is
frustratingly inaccurate, and this really erodes the benefit of
continuous integration.

> It would be handy if I knew a few things.  How to pass in VM debug args
> to Maven for its test runs, where to set the OpenJPA logging when Maven
> runs the tests, and how to pass in the appropriate parameters when
> running the test case outside of Maven.  Any insights on these questions
> will be appreciated.

I usually just add log configuration to the test cases themselves, in
the setUp(...) method:

    setUp(Class.class, Class2.class, ..., "openjpa.Log", "SQL=TRACE");

I believe that it is also possible to set this at the command line via
-Dopenjpa.Log=SQL=TRACE types of clauses.

-Patrick

On Jan 31, 2008 4:08 PM, David Ezzio <dezzio@apache.org> wrote:
> Hi,
>
> Progress has been slow on the test failures apparently introduced by
> 616972.  I have made progress in reproducing the results.
>
> The miscounting in TeamCity is gross.  Although it lists 33 failing
> tests complete with a credible stack trace for each, in fact, there is
> only one failing test.
>
> Meanwhile, I believe now, that change 616658 is no longer a suspect.  I
> get very erratic results here, but they seem to vary more by whether I
> run "mvn test" after "mvn clean" or "mvn install" after "mvn clean".  By
> doing just what TeamCity does, "mvn clean compile test", I can finally
> reproduce the results locally that TeamCity reports in its Compile-Derby
> build for this change (395, I think).  In spite of this, if I run "mvn
> -Dtest=TestCaseInsensitiveKeywordsInJPQL clean compile test" the test
> passes.
>
> It would be handy if I knew a few things.  How to pass in VM debug args
> to Maven for its test runs, where to set the OpenJPA logging when Maven
> runs the tests, and how to pass in the appropriate parameters when
> running the test case outside of Maven.  Any insights on these questions
> will be appreciated.
>
> But at least at this point, I can reproduce locally what TeamCity
> reports, and although it happened only once out of one try, I am
> optimistic that it is reproducible.  I will continue working this
> tomorrow.
>
> If the hangup is intolerable, let me know, and I'll back out my change
> tomorrow.
>
> David Ezzio
>
>
> David Ezzio wrote:
> > Hi,
> >
> > I am convinced that change 616658 (small change to DBDictionary.java) is
> > the source of the two intermittent failures that I am seeing locally.
> > One of those intermittent failures is the failure in
> > TestCaseInsensitiveKeywordsInJPQL.testCaseInsensitiveBooleans which is
> > one of the 33 related failures currently seen in TeamCity Derby builds.
> >
> > The other intermittent failure is TestVersion.testVersionTimestamp which
> > I reported earlier as occurring locally with change 616658.
> >
> > When I revert change 616658 locally, both of these issues go away even
> > with the changes that I submitted to AbstractBrokerFactory in change
> > 616972.
> >
> > I have a question: is it possible to run a build in TeamCity without
> > submitting the change (or conditionally submitting the change) to SVN?
> > I'd like to find out if reverting 616658 will fix the issue.  If so,
> > then I can try to figure out why it's a problem.
> >
> > Thanks,
> >
> > David
> >
> > David Ezzio wrote:
> >> No, but I'm taking a look since my changes appear to be the cause.
> >>
> >> Patrick Linskey wrote:
> >>> 5. The tests are now failing in the automated build; any idea why
> >>> this might be?
> >>>
> >>> -Patrick
> >>>
> >>> On Jan 30, 2008 6:11 PM, Patrick Linskey <plinskey@gmail.com> wrote:
> >>>> Hi,
> >>>>
> >>>> A few questions:
> >>>>
> >>>> 1. Were you seeing this cause problems somewhere? I.e., when do we
> >>>> serialize BrokerFactory instances?
> >>>>
> >>>> 2. Why not just move the initialization code to the declaration line
> >>>> (instead of the constructor)?
> >>>>
> >>>> 3. It looks like there are other transactional structures that might
> >>>> not be properly handled in serialization / deserialization (lifecycle
> >>>> listeners, for example). Should we be making more changes there?
> >>>>
> >>>> -Patrick
> >>>>
> >>>>
> >>>> On Jan 30, 2008 4:59 PM,  <dezzio@apache.org> wrote:
> >>>>> Author: dezzio
> >>>>> Date: Wed Jan 30 16:59:02 2008
> >>>>> New Revision: 616972
> >>>>>
> >>>>> URL: http://svn.apache.org/viewvc?rev=616972&view=rev
> >>>>> Log:
> >>>>> Allow EntityManagerFactory objects to be serialized and
> >>>>> deserialized successfully.
> >>>>>
> >>>>> Added:
> >>>>>
> >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java
> >>>>> (with props)
> >>>>> Modified:
> >>>>>
> >>>>> openjpa/trunk/openjpa-kernel/src/main/java/org/apache/openjpa/kernel/AbstractBrokerFactory.java
> >>>>>
> >>>>>
> >>>>> Modified:
> >>>>> openjpa/trunk/openjpa-kernel/src/main/java/org/apache/openjpa/kernel/AbstractBrokerFactory.java
> >>>>>
> >>>>> URL:
> >>>>> http://svn.apache.org/viewvc/openjpa/trunk/openjpa-kernel/src/main/java/org/apache/openjpa/kernel/AbstractBrokerFactory.java?rev=616972&r1=616971&r2=616972&view=diff
> >>>>>
> >>>>> ==============================================================================
> >>>>>
> >>>>> ---
> >>>>> openjpa/trunk/openjpa-kernel/src/main/java/org/apache/openjpa/kernel/AbstractBrokerFactory.java
> >>>>> (original)
> >>>>> +++
> >>>>> openjpa/trunk/openjpa-kernel/src/main/java/org/apache/openjpa/kernel/AbstractBrokerFactory.java
> >>>>> Wed Jan 30 16:59:02 2008
> >>>>> @@ -147,8 +147,7 @@
> >>>>>       */
> >>>>>      protected AbstractBrokerFactory(OpenJPAConfiguration config)
{
> >>>>>          _conf = config;
> >>>>> -        _pcClassLoaders = new ConcurrentReferenceHashSet(
> >>>>> -            ConcurrentReferenceHashSet.WEAK);
> >>>>> +        getPcClassLoaders();
> >>>>>      }
> >>>>>
> >>>>>      /**
> >>>>> @@ -287,13 +286,13 @@
> >>>>>                      if (needsSub(cls))
> >>>>>                          toRedefine.add(cls);
> >>>>>                  }
> >>>>> -                _pcClassLoaders.add(loader);
> >>>>> +                getPcClassLoaders().add(loader);
> >>>>>                  _pcClassNames = c;
> >>>>>              }
> >>>>>              _persistentTypesLoaded = true;
> >>>>>          } else {
> >>>>>              // reload with this loader
> >>>>> -            if (_pcClassLoaders.add(loader)) {
> >>>>> +            if (getPcClassLoaders().add(loader)) {
> >>>>>                  for (Iterator itr = _pcClassNames.iterator();
> >>>>> itr.hasNext();) {
> >>>>>                      try {
> >>>>>                          Class cls =
> >>>>> @@ -818,4 +817,15 @@
> >>>>>              _transactional.remove (_trans);
> >>>>>                 }
> >>>>>         }
> >>>>> +
> >>>>> +   /**
> >>>>> +    * Method insures that deserialized EMF has this reference
> >>>>> re-instantiated
> >>>>> +    */
> >>>>> +   private Collection getPcClassLoaders() {
> >>>>> +      if (_pcClassLoaders == null)
> >>>>> +        _pcClassLoaders = new ConcurrentReferenceHashSet(
> >>>>> +            ConcurrentReferenceHashSet.WEAK);
> >>>>> +
> >>>>> +      return _pcClassLoaders;
> >>>>> +   }
> >>>>>  }
> >>>>>
> >>>>> Added:
> >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java
> >>>>>
> >>>>> URL:
> >>>>> http://svn.apache.org/viewvc/openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java?rev=616972&view=auto
> >>>>>
> >>>>> ==============================================================================
> >>>>>
> >>>>> ---
> >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java
> >>>>> (added)
> >>>>> +++
> >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java
> >>>>> Wed Jan 30 16:59:02 2008
> >>>>> @@ -0,0 +1,74 @@
> >>>>> +/*
> >>>>> + * Licensed to the Apache Software Foundation (ASF) under one
> >>>>> + * or more contributor license agreements.  See the NOTICE file
> >>>>> + * distributed with this work for additional information
> >>>>> + * regarding copyright ownership.  The ASF licenses this file
> >>>>> + * to you under the Apache License, Version 2.0 (the
> >>>>> + * "License"); you may not use this file except in compliance
> >>>>> + * with the License.  You may obtain a copy of the License at
> >>>>> + *
> >>>>> + * http://www.apache.org/licenses/LICENSE-2.0
> >>>>> + *
> >>>>> + * Unless required by applicable law or agreed to in writing,
> >>>>> + * software distributed under the License is distributed on an
> >>>>> + * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
> >>>>> + * KIND, either express or implied.  See the License for the
> >>>>> + * specific language governing permissions and limitations
> >>>>> + * under the License.
> >>>>> + */
> >>>>> +package org.apache.openjpa.persistence.simple;
> >>>>> +
> >>>>> +import java.io.*;
> >>>>> +
> >>>>> +import javax.persistence.EntityManager;
> >>>>> +import javax.persistence.EntityManagerFactory;
> >>>>> +import javax.persistence.EntityTransaction;
> >>>>> +
> >>>>> +import junit.textui.TestRunner;
> >>>>> +import org.apache.openjpa.persistence.OpenJPAEntityManager;
> >>>>> +import org.apache.openjpa.persistence.test.SingleEMFTestCase;
> >>>>> +
> >>>>> +/**
> >>>>> + * Tests that a EntityManagerFactory can be used after serialization.
> >>>>> + *
> >>>>> + * @author David Ezzio
> >>>>> + */
> >>>>> +public class TestSerializedFactory
> >>>>> +    extends SingleEMFTestCase {
> >>>>> +
> >>>>> +    public void setUp() {
> >>>>> +        setUp(AllFieldTypes.class);
> >>>>> +    }
> >>>>> +
> >>>>> +    public void testSerializedEntityManagerFactory() throws
> >>>>> Exception {
> >>>>> +        // serialize and deserialize the entity manager factory
> >>>>> +        ByteArrayOutputStream baos = new ByteArrayOutputStream();
> >>>>> +        ObjectOutputStream oos = new ObjectOutputStream(baos);
> >>>>> +        oos.writeObject(emf);
> >>>>> +        EntityManagerFactory emf2 =
> >>>>> +            (EntityManagerFactory) new ObjectInputStream(
> >>>>> +            new
> >>>>> ByteArrayInputStream(baos.toByteArray())).readObject();
> >>>>> +
> >>>>> +        // use the deserialized entity manager factory
> >>>>> +        assertTrue("The deserialized entity manager factory is
not
> >>>>> open",
> >>>>> +            emf2.isOpen());
> >>>>> +        EntityManager em = emf2.createEntityManager();
> >>>>> +        assertTrue("The newly created entity manager is not open",
> >>>>> em.isOpen());
> >>>>> +
> >>>>> +        // exercise the entity manager produced from the
> >>>>> deserialized EMF
> >>>>> +        em.getTransaction().begin();
> >>>>> +        em.persist(new AllFieldTypes());
> >>>>> +        em.getTransaction().commit();
> >>>>> +
> >>>>> +        // close the extra resources
> >>>>> +        em.close();
> >>>>> +        assertFalse("The entity manager is not closed", em.isOpen());
> >>>>> +        emf2.close();
> >>>>> +        assertFalse("The entity manager factory is not closed",
> >>>>> emf2.isOpen());
> >>>>> +    }
> >>>>> +
> >>>>> +    public static void main(String[] args) {
> >>>>> +        TestRunner.run(TestSerializedFactory.class);
> >>>>> +    }
> >>>>> +}
> >>>>> +
> >>>>>
> >>>>> Propchange:
> >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apache/openjpa/persistence/simple/TestSerializedFactory.java
> >>>>>
> >>>>> ------------------------------------------------------------------------------
> >>>>>
> >>>>>     svn:eol-style = native
> >>>>>
> >>>>>
> >>>>>
> >>>>
> >>>>
> >>>> --
> >>>> Patrick Linskey
> >>>> 202 669 5907
> >>>>
> >>>
> >>>
> >>>
> >>
> >
>



-- 
Patrick Linskey
202 669 5907

Mime
View raw message