openjpa-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Kevin Sutter" <kwsut...@gmail.com>
Subject Re: Failures in TeamCity Derby build and svn commit: r616972
Date Fri, 01 Feb 2008 17:10:33 GMT
On Feb 1, 2008 10:17 AM, Pinaki Poddar <ppoddar@bea.com> wrote:

> Factories needs to be Serializable as they are often placed in JNDI.


The JNDI lookup needs to work for the Factory definition, but it doesn't
require that the actual Factory be present in JNDI.  And, since the JNDI
lookup is meant for the EE (Container) environment, I would guess that most
implementations provide a wrapper object in JNDI for the actual Factory.

Granted, this wrapper approach is not a requirement.  But, I can't find the
requirement in the spec that the EMF has to be Serializable either...  :-)

Kevin


>
> -----Original Message-----
> From: Patrick Linskey [mailto:plinskey@gmail.com]
> Sent: Thursday, January 31, 2008 7:57 PM
> To: dev@openjpa.apache.org
> Subject: Re: Failures in TeamCity Derby build and svn commit: r616972
>
> To be honest, I don't really understand why a BrokerFactory needs to be
> serializable, so I think probably it'd be good to make sure that there
> is good reason for that before going too deep into this debugging
> exercise.
>
> -Patrick
>
> On Jan 31, 2008 5:52 PM, Patrick Linskey <plinskey@gmail.com> wrote:
> > > 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/apach
> > > >>>>> e/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/apach
> > > >>>>> e/openjpa/persistence/simple/TestSerializedFactory.java
> > > >>>>>
> > > >>>>> URL:
> > > >>>>> http://svn.apache.org/viewvc/openjpa/trunk/openjpa-persistence
> > > >>>>> -jdbc/src/test/java/org/apache/openjpa/persistence/simple/Test
> > > >>>>> SerializedFactory.java?rev=616972&view=auto
> > > >>>>>
> > > >>>>> ==============================================================
> > > >>>>> ================
> > > >>>>>
> > > >>>>> ---
> > > >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apach
> > > >>>>> e/openjpa/persistence/simple/TestSerializedFactory.java
> > > >>>>> (added)
> > > >>>>> +++
> > > >>>>> openjpa/trunk/openjpa-persistence-jdbc/src/test/java/org/apach
> > > >>>>> e/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/apach
> > > >>>>> e/openjpa/persistence/simple/TestSerializedFactory.java
> > > >>>>>
> > > >>>>> --------------------------------------------------------------
> > > >>>>> ----------------
> > > >>>>>
> > > >>>>>     svn:eol-style = native
> > > >>>>>
> > > >>>>>
> > > >>>>>
> > > >>>>
> > > >>>>
> > > >>>> --
> > > >>>> Patrick Linskey
> > > >>>> 202 669 5907
> > > >>>>
> > > >>>
> > > >>>
> > > >>>
> > > >>
> > > >
> > >
> >
> >
> >
> > --
> > Patrick Linskey
> > 202 669 5907
> >
>
>
>
> --
> Patrick Linskey
> 202 669 5907
>
> Notice:  This email message, together with any attachments, may contain
> information  of  BEA Systems,  Inc.,  its subsidiaries  and  affiliated
> entities,  that may be confidential,  proprietary,  copyrighted  and/or
> legally privileged, and is intended solely for the use of the individual or
> entity named in this message. If you are not the intended recipient, and
> have received this message in error, please immediately return this by email
> and then delete it.
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message