o.j.c.statetransfer.VersionedTestBase
by Manik Surtani
All,
Any idea why VersionedTestBase.testPartialStateTferWithClassLoader()
is implemented the way it is?
Line 293 has:
assertNull("/a/b not transferred to loader", loader.get(A_B));
and soon after that (no more mods to the caches except some
assertions) on line 302:
assertEquals("Correct state from loader for /a/b", ben.toString(),
loader.get(A_B).get("person").toString());
This basically throws an NPE on loader.get(A_B).get("person") but
this NPE is *expected*, based on the assertion before... ? Am I
missing something here?
Cheers,
--
Manik Surtani
Lead, JBoss Cache
JBoss, a division of Red Hat
Email: manik(a)jboss.org
Telephone: +44 7786 702 706
MSN: manik(a)surtani.org
Yahoo/AIM/Skype: maniksurtani
18 years, 2 months
RE: [jbosscache-dev] JBoss Cache 2.0.0.ALPHA status
by Brian Stansberry
Vladimir Blagojevic wrote:
> I made TcpCacheLoader test pass but tcpcacheloader will not
> propagate load/store. This was the case even before I made
> the cacheloader change.
> Similar situation is with rpc and rmi loaders and we have to
> figure out what to do there.
In the case of store, is this a matter of reading out the NodeData
objects and invoking a put(Fqn, Map) on the remote cache for each one?
For load, a bunch of recursive get(Fqn) calls to create the NodeData?
(I'm skipping any subtleties of the 2.0 API here.)
A problem with the put(Fqn, Map) approach is that call integrates the
passed Map into the target node, rather than replacing the map contents.
So you could be left with stale key/value pairs in the far cache.
18 years, 2 months
RE: JBC cruise control stuff
by Rajesh Rajasekaran
I could list modifications from last build rather than modifications
from last successful build.
Also the modification list would display the date and time of each
modification. This might be helpful in tracking any changes.
Also there should be a way where we could list all modifications without
the build emails running too big. (Currently we restrict to 50
modifications)
We are looking at an upgrade.
-----Original Message-----
From: Manik Surtani [mailto:manik@jboss.org]
Sent: Monday, September 11, 2006 10:38 AM
To: Rajesh Rajasekaran
Cc: Vladimir Blagojevic; jbosscache-dev(a)lists.jboss.org; QA
Subject: Re: JBC cruise control stuff
On 11 Sep 2006, at 16:33, Rajesh Rajasekaran wrote:
> The JBossCache testsuite doesn't run on a separate server. It was
> moved
> to a server with a lighter load, no long testsuites running, so we can
> have a guaranteed build every day. There are other builds running on
> that server 24/7.
>
> I don't have much control with the current version of CC, the
> modification list gets accumulated with every build until we have a
> successful build.
What do newer versions of CC have to offer in this regard, and is it
feasible to upgrade if necessary?
> I can add a modification.log to every build which lists only the
> modifications for that particular build, which runs nightly. Hope
> that'll be helpful.
>
>
>
> -----Original Message-----
> From: Manik Surtani [mailto:manik@jboss.org]
> Sent: Monday, September 11, 2006 10:20 AM
> To: Vladimir Blagojevic
> Cc: jbosscache-dev(a)lists.jboss.org; QA
> Subject: Re: JBC cruise control stuff
>
> Hmm, I wonder. I doubt it though, since CC does nightly check outs
> and runs the test suite rather than whenever a commit is made. I
> presume this has to do with how long the test suite takes to run (80
> mins at the moment), but I still see a lot of merit in having the CC
> test suite runs triggered by check ins.
>
> QA, is this feasible now, since the JBC test suite is now run on a
> separate server?
>
> Cheers,
> --
> Manik Surtani
>
> Lead, JBoss Cache
> JBoss, a division of Red Hat
>
> Email: manik(a)jboss.org
> Telephone: +44 7786 702 706
> MSN: manik(a)surtani.org
> Yahoo/AIM/Skype: maniksurtani
>
>
> On 11 Sep 2006, at 16:11, Vladimir Blagojevic wrote:
>
>> Hey Manik,
>>
>> Is there a way to cross reference list of cvs modifications with test
>> results? For example, it would be cool if I can browse history of
>> state
>> transfer results and detect which modifications caused some of the
>> state
>> transfer tests to fail.
>>
>> Peace,
>> Vladimir
>
18 years, 2 months
RE: JBC cruise control stuff
by Rajesh Rajasekaran
The JBossCache testsuite doesn't run on a separate server. It was moved
to a server with a lighter load, no long testsuites running, so we can
have a guaranteed build every day. There are other builds running on
that server 24/7.
I don't have much control with the current version of CC, the
modification list gets accumulated with every build until we have a
successful build.
I can add a modification.log to every build which lists only the
modifications for that particular build, which runs nightly. Hope
that'll be helpful.
-----Original Message-----
From: Manik Surtani [mailto:manik@jboss.org]
Sent: Monday, September 11, 2006 10:20 AM
To: Vladimir Blagojevic
Cc: jbosscache-dev(a)lists.jboss.org; QA
Subject: Re: JBC cruise control stuff
Hmm, I wonder. I doubt it though, since CC does nightly check outs
and runs the test suite rather than whenever a commit is made. I
presume this has to do with how long the test suite takes to run (80
mins at the moment), but I still see a lot of merit in having the CC
test suite runs triggered by check ins.
QA, is this feasible now, since the JBC test suite is now run on a
separate server?
Cheers,
--
Manik Surtani
Lead, JBoss Cache
JBoss, a division of Red Hat
Email: manik(a)jboss.org
Telephone: +44 7786 702 706
MSN: manik(a)surtani.org
Yahoo/AIM/Skype: maniksurtani
On 11 Sep 2006, at 16:11, Vladimir Blagojevic wrote:
> Hey Manik,
>
> Is there a way to cross reference list of cvs modifications with test
> results? For example, it would be cool if I can browse history of
> state
> transfer results and detect which modifications caused some of the
> state
> transfer tests to fail.
>
> Peace,
> Vladimir
18 years, 2 months
Re: JBC cruise control stuff
by Manik Surtani
Hmm, I wonder. I doubt it though, since CC does nightly check outs
and runs the test suite rather than whenever a commit is made. I
presume this has to do with how long the test suite takes to run (80
mins at the moment), but I still see a lot of merit in having the CC
test suite runs triggered by check ins.
QA, is this feasible now, since the JBC test suite is now run on a
separate server?
Cheers,
--
Manik Surtani
Lead, JBoss Cache
JBoss, a division of Red Hat
Email: manik(a)jboss.org
Telephone: +44 7786 702 706
MSN: manik(a)surtani.org
Yahoo/AIM/Skype: maniksurtani
On 11 Sep 2006, at 16:11, Vladimir Blagojevic wrote:
> Hey Manik,
>
> Is there a way to cross reference list of cvs modifications with test
> results? For example, it would be cool if I can browse history of
> state
> transfer results and detect which modifications caused some of the
> state
> transfer tests to fail.
>
> Peace,
> Vladimir
18 years, 2 months
Re: Nice feature in JBossCache
by Manik Surtani
Sounds good - are we thinking for 2.0.0? If this is urgent, this
could even be added to the 1.4.0 branch.
Cheers,
--
Manik Surtani
Lead, JBoss Cache
JBoss, a division of Red Hat
Email: manik(a)jboss.org
Telephone: +44 7786 702 706
MSN: manik(a)surtani.org
Yahoo/AIM/Skype: maniksurtani
On 9 Sep 2006, at 15:17, Bela Ban wrote:
> would be to have an operation (exposed by JMX) which reports the
> last N view changing operations, e.g. member joined, left, was
> suspected etc. This is part of ClusterPartition, and it really
> helped me debugging some issues, especially if the log files are huge.
>
> String showHistory()
> --
> Bela Ban
> Lead JGroups / Manager JBoss Clustering Group
> JBoss - a division of Red Hat
18 years, 2 months
Re: Where is org.jnp.interfaces.Naming ?
by Manik Surtani
Was this from a clean checkout?
This test ran just fine on last night's CC run:
http://cruisecontrol.jboss.com/cc/artifacts/jboss-cache-testsuite/
20060908231047/results/index.html
Runs fine for me too, in IntelliJ.
Cheers,
--
Manik Surtani
Lead, JBoss Cache
JBoss, a division of Red Hat
Email: manik(a)jboss.org
Telephone: +44 7786 702 706
MSN: manik(a)surtani.org
Yahoo/AIM/Skype: maniksurtani
On 9 Sep 2006, at 12:48, Bela Ban wrote:
> I wanted to run a simple unit test (JBossCache CVS head), the
> missing class is in none of the JARs in lib
>
> java.lang.NoClassDefFoundError: org/jnp/interfaces/Naming
> at java.lang.Class.forName0(Native Method)
> at java.lang.Class.forName(Class.java:242)
> at com.sun.naming.internal.VersionHelper12.loadClass
> (VersionHelper12.java:50)
> at com.sun.naming.internal.ResourceManager.getFactory
> (ResourceManager.java:371)
> at javax.naming.spi.NamingManager.getURLObject
> (NamingManager.java:575)
> at javax.naming.spi.NamingManager.getURLContext
> (NamingManager.java:533)
> at javax.naming.InitialContext.getURLOrDefaultInitCtx
> (InitialContext.java:279)
> at javax.naming.InitialContext.bind(InitialContext.java:359)
> at
> org.jboss.cache.transaction.DummyTransactionManager.getInstance
> (DummyTransactionManager.java:33)
> at org.jboss.cache.replicated.SyncReplTxTest.setUp
> (SyncReplTxTest.java:81)
> --
> Bela Ban
> Lead JGroups / Manager JBoss Clustering Group
> JBoss - a division of Red Hat
18 years, 2 months
Re: JBossCache CVS head
by Manik Surtani
On 9 Sep 2006, at 13:13, Bela Ban wrote:
> Okay,
> I'm checking out a new version, compile it (./build.sh) and then
> call ./build.sh functionaltests. A lot of tests fail because that
> class I mentioned previously is not found.
Again, from the CC run, we've got over 90% pass rate with a few
integrations still pending:
http://cruisecontrol.jboss.com/cc/artifacts/jboss-cache-testsuite/
20060908231047/results/index.html
Let me run them on the CMD line here and see if there is some
classpath issue perhaps with the build script.
>
> Also, why am I compiling the AOP examples if all I want to do is to
> run the functional tests ? There is an error message related to AOP
> too (see below).
>
> Also, why do we have the aopc compiler spit out messages at the (so
> it seems) TRACE level ? Can we make this less verbose ?
>
> Is JBossCache CVS head currently broken ? Otherwise I would expect
> that the sequence CVS checkout, compile, run testsuite always works !
It should not be. If it is, it is probably due to the jars being
moved around/upgraded over the last few days, but I would have seen
this come up sooner.
>
>
> bela@dell /cygdrive/c/JBossCache
> $ ./build.sh functionaltests
> Buildfile: build.xml
>
> compile-classes14:
> [javac] Compiling 1 source file to C:\JBossCache\output\test-
> classes
> [copy] Copying 38 files to C:\JBossCache\output\etc
> [copy] Copying 1 file to C:\JBossCache\output\resources
>
> compile-classes50-1:
>
> aopc-pojocache:
> [aopc] [info] Total length of filenames to be compiled is
> greater than 1000, listing files in --SOURCEPATH: c:\DOCUME~1
> \BELA~1.DEL\LOCALS~1\Temp\src49385.tmp
> [aopc] [debug] Passed in instrumentor: null
> [aopc] [debug] Defaulting instrumentor to:
> org.jboss.aop.instrument.ClassicInstrumentor
> [aopc] [debug] Passed in advisor: null
> [aopc] [debug] Defaulting advisor to: org.jboss.aop.ClassAdvisor
> [aopc] [debug] jboss.aop.class.path is NULL
> [aopc] [debug] jboss.aop.search.classpath: 'null' true
> [aopc] [debug] jboss.aop.path: C:\JBossCache\output\resources
> \jboss-aop.xml
>
> --
> Bela Ban
> Lead JGroups / Manager JBoss Clustering Group
> JBoss - a division of Red Hat
18 years, 2 months
Re: Questions and suggestions regarding JBossCache 2.0
by Manik Surtani
On 9 Sep 2006, at 12:46, Bela Ban wrote:
> Why isn't there a CacheFactory.createCache() (using the default
> configuration)
> I think it would be useful to have a static method in Configuration
> which returns the default config, so we could do: Cache c=new
> DefaultCacheFactory().createCache(Configuration.getDefault())
> Configuration should have a copy() or clone() method, so we can
> reuse a config, and make small modifications
> Make Configuration serializable (ship them around ?)
>
All good suggestions. +1 for all of these.
Any suggestions on what should be in the default configuration, apart
from LOCAL/PESS. LOCKING/REPEATABLE READ and no cacheloaders/eviction
policies?
18 years, 2 months
TcpCacheLoader cannot support load/store state operations
by Vladimir Blagojevic
Hi,
We are not going to be able to support load/store state operations with
TcpCacheLoader so easily since we can not serialize stream handles to
remote delegate cache. I will leave load/store operations throwing
UnsupportedOperationException and TcpCacheLoaderTest is an expected
failure. It seems like it will be a similar case with rpc and rmi based
cacheloaders. We have to think of a solution for these cases that will
not be a kludge.
Vladimir
18 years, 2 months
