p.s. I'm currently using http://watchout4snakes.com/creativitytools/RandomWord/RandomWordPlus.aspx as random word generator. On 09/11/2009 12:12 PM, Galder Zamarreno wrote: > Guys, > > Looking at the Infinispan tests, I've seen plenty same keys/values > repeated in loads of tests, i.e. k1, k2, v1, v2....etc, so sometimes, > when you're debugging a specific tests, particularly when running the > whole testsuite, it's hard to pint point the data related to your > test > since same key/values are used by several tests. > > So, to try to help with finding the data related to my tests, I'm > trying > to use random word generator to generate uncommon or obscure words > that > I can use in the tests. Doing this allows me to quickly find my > data in > the logs, i.e. "key-isoprene" > > I don't know whether other people are interesting in using a similar > method but at least for me, it helps. > > Cheers, -- Galder Zamarreño Sr. Software Engineer Infinispan, JBoss Cache _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

On 09/11/2009 02:23 PM, Mircea Markus wrote: > Another thing that helps is using the thread name. It's true that > there are some threads created within jgroups, and these won't match > but generally it helps. > Before starting a test, a log message is being output - this contains > the name of the test that is to be executed. Same after the test is > finished. Logs look like: > "Starting test..." > and > "Test Xyz failed"/"Test Xyz skipped"/"Test Xyz succeeded" True, those tips help as well. The easier is to determine the log associated with a failing test, the easier debugging will be. I've also been thinking of a way to improve logging based on the Hibernate 2nd level cache work I did. It'd be interesting that any instance of a class that is specific to a cache instanc, i.e. "galder", had the cache name appended.

In environments where there're several cache instances deployed, i.e. AS, this would help greatly. Example: Instead of: 2009-09-11 16:41:05,635 591 TRACE [org.infinispan.interceptors.InvocationContextInterceptor] (main:) Invoked with command PutKeyValueCommand{key=key, value=value, putIfAbsent=false, lifespanMillis=-1, maxIdleTimeMillis=-1} and InvocationContext [NonTxInvocationContext{flags=null}] ... 2009-09-11 16:41:05,635 591 TRACE [org.infinispan.interceptors.CallInterceptor] (main:) Executing command: PutKeyValueCommand{key=key, value=value, putIfAbsent=false, lifespanMillis=-1, maxIdleTimeMillis=-1}. Something like: 2009-09-11 16:41:05,635 591 TRACE [org.infinispan.interceptors.InvocationContextInterceptor.galder] (main:) Invoked with command PutKeyValueCommand{key=key, value=value, putIfAbsent=false, lifespanMillis=-1, maxIdleTimeMillis=-1} and InvocationContext [NonTxInvocationContext{flags=null}] ... 2009-09-11 16:41:05,635 591 TRACE [org.infinispan.interceptors.CallInterceptor.galder] (main:) Executing command: PutKeyValueCommand{key=key, value=value, putIfAbsent=false, lifespanMillis=-1, maxIdleTimeMillis=-1}. Remember that by default, in Hibernate we create a cache instance per entity, so imagine the number of diff caches you might have to deal with in an app server env. > > On Sep 11, 2009, at 1:13 PM, Galder Zamarreno wrote: > >> p.s. I'm currently using >> http://watchout4snakes.com/creativitytools/RandomWord/RandomWordPlus.aspx >> as >> random word generator. >> >> On 09/11/2009 12:12 PM, Galder Zamarreno wrote: >>> Guys, >>> >>> Looking at the Infinispan tests, I've seen plenty same keys/values >>> repeated in loads of tests, i.e. k1, k2, v1, v2....etc, so >>> sometimes, >>> when you're debugging a specific tests, particularly when running >>> the >>> whole testsuite, it's hard to pint point the data related to your >>> test >>> since same key/values are used by several tests. >>> >>> So, to try to help with finding the data related to my tests, I'm >>> trying >>> to use random word generator to generate uncommon or obscure words >>> that >>> I can use in the tests. Doing this allows me to quickly find my >>> data in >>> the logs, i.e. "key-isoprene" >>> >>> I don't know whether other people are interesting in using a >>> similar >>> method but at least for me, it helps. >>> >>> Cheers, >> >> -- >> Galder Zamarreño >> Sr. Software Engineer >> Infinispan, JBoss Cache >> _______________________________________________ >> infinispan-dev mailing list >> infinispan-dev(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/infinispan-dev > > > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev -- Galder Zamarreño Sr. Software Engineer Infinispan, JBoss Cache _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

Another thing that helps is using the thread name. It's true that there are some threads created within jgroups, and these won't match but generally it helps. Before starting a test, a log message is being output - this contains the name of the test that is to be executed. Same after the test is finished. Logs look like: "Starting test..." and "Test Xyz failed"/"Test Xyz skipped"/"Test Xyz succeeded" On Sep 11, 2009, at 1:13 PM, Galder Zamarreno wrote: > p.s. I'm currently using > http://watchout4snakes.com/creativitytools/RandomWord/RandomWordPlus.aspx >  as > random word generator. > > On 09/11/2009 12:12 PM, Galder Zamarreno wrote: >> Guys, >> >> Looking at the Infinispan tests, I've seen plenty same keys/values >> repeated in loads of tests, i.e. k1, k2, v1, v2....etc, so sometimes, >> when you're debugging a specific tests, particularly when running the >> whole testsuite, it's hard to pint point the data related to your >> test >> since same key/values are used by several tests. >> >> So, to try to help with finding the data related to my tests, I'm >> trying >> to use random word generator to generate uncommon or obscure words >> that >> I can use in the tests. Doing this allows me to quickly find my >> data in >> the logs, i.e. "key-isoprene" >> >> I don't know whether other people are interesting in using a similar >> method but at least for me, it helps. >> >> Cheers, > > -- > Galder Zamarreño > Sr. Software Engineer > Infinispan, JBoss Cache > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

A crazier alternative is to grab the testMethod name and use that as part of the key/string - you can do it by getting the stack trace and a little bit of work (I had a helper method somewhere). Something like: Thread.currentThread().getStackTrace()(1).getMethodName() (at the top of the test method)-- its probably 1 as 0 would be the getStackTrace method itself... I have found it helpful when digging through things and I see a value from a certain test...

On Fri, Sep 11, 2009 at 10:23 PM, Mircea Markus<mircea.markus(a)jboss.com> wrote: > Another thing that helps is using the thread name. It's true that > there are some threads created within jgroups, and these won't match > but generally it helps. > Before starting a test, a log message is being output - this contains > the name of the test that is to be executed. Same after the test is > finished. Logs look like: > "Starting test..." > and > "Test Xyz failed"/"Test Xyz skipped"/"Test Xyz succeeded" > > On Sep 11, 2009, at 1:13 PM, Galder Zamarreno wrote: > >> p.s. I'm currently using >> http://watchout4snakes.com/creativitytools/RandomWord/RandomWordPlus.aspx >> as >> random word generator. >> >> On 09/11/2009 12:12 PM, Galder Zamarreno wrote: >>> Guys, >>> >>> Looking at the Infinispan tests, I've seen plenty same keys/values >>> repeated in loads of tests, i.e. k1, k2, v1, v2....etc, so sometimes, >>> when you're debugging a specific tests, particularly when running the >>> whole testsuite, it's hard to pint point the data related to your >>> test >>> since same key/values are used by several tests. >>> >>> So, to try to help with finding the data related to my tests, I'm >>> trying >>> to use random word generator to generate uncommon or obscure words >>> that >>> I can use in the tests. Doing this allows me to quickly find my >>> data in >>> the logs, i.e. "key-isoprene" >>> >>> I don't know whether other people are interesting in using a similar >>> method but at least for me, it helps. >>> >>> Cheers, >> >> -- >> Galder Zamarreño >> Sr. Software Engineer >> Infinispan, JBoss Cache >> _______________________________________________ >> infinispan-dev mailing list >> infinispan-dev(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/infinispan-dev > > > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev >

On Sep 14, 2009, at 12:02 PM, Galder Zamarreno wrote: > > > On 09/12/2009 12:31 AM, Michael Neale wrote: >> A crazier alternative is to grab the testMethod name and use that as >> part of the key/string - you can do it by getting the stack trace and >> a little bit of work (I had a helper method somewhere). >> >> Something like: >> Thread.currentThread().getStackTrace()(1).getMethodName() (at the top >> of the test method)-- its probably 1 as 0 would be the getStackTrace >> method itself... I have found it helpful when digging through things >> and I see a value from a certain test... That's a good idea but doesn't apply when there's inheritance between test classes (and ISPN uses inheritance). Eg. abstract class Base { public void testMethod(){} } @Test public class A extends Base{} @Test public class B extends Base{} Now test method will be run twice, and the stack trace won't let you make the difference between it being run in the context of class A or of class B. But yep, in all other scenarios this is a very useful mechanism.

> > Maybe you could add this debugging method as a static method to > TestingUtil class? > > Btw, I've updated > http://www.jboss.org/community/wiki/InfinispanQuickStartGuideforDevelopers > with these two tips. > >> >> >> On Fri, Sep 11, 2009 at 10:23 PM, Mircea Markus<mircea.markus(a)jboss.com >>> wrote: >>> Another thing that helps is using the thread name. It's true that >>> there are some threads created within jgroups, and these won't match >>> but generally it helps. >>> Before starting a test, a log message is being output - this >>> contains >>> the name of the test that is to be executed. Same after the test is >>> finished. Logs look like: >>> "Starting test..." >>> and >>> "Test Xyz failed"/"Test Xyz skipped"/"Test Xyz succeeded" >>> >>> On Sep 11, 2009, at 1:13 PM, Galder Zamarreno wrote: >>> >>>> p.s. I'm currently using >>>> http://watchout4snakes.com/creativitytools/RandomWord/RandomWordPlus.aspx >>>> as >>>> random word generator. >>>> >>>> On 09/11/2009 12:12 PM, Galder Zamarreno wrote: >>>>> Guys, >>>>> >>>>> Looking at the Infinispan tests, I've seen plenty same keys/values >>>>> repeated in loads of tests, i.e. k1, k2, v1, v2....etc, so >>>>> sometimes, >>>>> when you're debugging a specific tests, particularly when >>>>> running the >>>>> whole testsuite, it's hard to pint point the data related to your >>>>> test >>>>> since same key/values are used by several tests. >>>>> >>>>> So, to try to help with finding the data related to my tests, I'm >>>>> trying >>>>> to use random word generator to generate uncommon or obscure words >>>>> that >>>>> I can use in the tests. Doing this allows me to quickly find my >>>>> data in >>>>> the logs, i.e. "key-isoprene" >>>>> >>>>> I don't know whether other people are interesting in using a >>>>> similar >>>>> method but at least for me, it helps. >>>>> >>>>> Cheers, >>>> >>>> -- >>>> Galder Zamarreño >>>> Sr. Software Engineer >>>> Infinispan, JBoss Cache >>>> _______________________________________________ >>>> infinispan-dev mailing list >>>> infinispan-dev(a)lists.jboss.org >>>> https://lists.jboss.org/mailman/listinfo/infinispan-dev >>> >>> >>> _______________________________________________ >>> infinispan-dev mailing list >>> infinispan-dev(a)lists.jboss.org >>> https://lists.jboss.org/mailman/listinfo/infinispan-dev >>> >> >> >> > > -- > Galder Zamarreño > Sr. Software Engineer > Infinispan, JBoss Cache > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

On 09/15/2009 09:26 AM, Mircea Markus wrote: > > On Sep 14, 2009, at 12:02 PM, Galder Zamarreno wrote: > >> >> >> On 09/12/2009 12:31 AM, Michael Neale wrote: >>> A crazier alternative is to grab the testMethod name and use that as >>> part of the key/string - you can do it by getting the stack trace and >>> a little bit of work (I had a helper method somewhere). >>> >>> Something like: >>> Thread.currentThread().getStackTrace()(1).getMethodName() (at the top >>> of the test method)-- its probably 1 as 0 would be the getStackTrace >>> method itself... I have found it helpful when digging through things >>> and I see a value from a certain test...

That's nice! Does it work with inheritance? I.e. Is there a way to determine the runtime class (vs defining class) of the given Method object?

On Sep 15, 2009, at 1:31 PM, Galder Zamarreno wrote: > testng will populate it with the test method information _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

sweet, thanks for the tip. In JbossCache I did an uglier workaround to fix this, passed the Class object to the factory that supplied CacheManagers, but this approach is much nicer.

On Sep 15, 2009, at 1:45 PM, Galder Zamarreno wrote: > > > On 09/15/2009 12:35 PM, Mircea Markus wrote: >> That's nice! >> Does it work with inheritance? I.e. Is there a way to determine the >> runtime class (vs defining class) of the given Method object? > > I think that if you pass an ITestContext, using the getName() > method, u > can get the runtime class. > > Example: > > CacheAPITest: > > public void testGetMembersInLocalMode(ITestContext context) { > System.out.println("Booooooooooo moooooooo " + > context.getName()); > assert cache.getCacheManager().getAddress() == null : "Cache > members should be null if running in LOCAL mode"; > } > > When I run CacheAPIMVCCTest, it prints: > > Booooooooooo moooooooo > org.infinispan.api.mvcc.repeatable_read.CacheAPIMVCCTest > > So that way, u can get the runtime class name. > >> On Sep 15, 2009, at 1:31 PM, Galder Zamarreno wrote: >> >>> testng will populate it with the test method information >> >> _______________________________________________ >> infinispan-dev mailing list >> infinispan-dev(a)lists.jboss.org >> https://lists.jboss.org/mailman/listinfo/infinispan-dev > > -- > Galder Zamarreño > Sr. Software Engineer > Infinispan, JBoss Cache > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

On 09/15/2009 12:14 PM, Galder Zamarreno wrote: > > > On 09/15/2009 09:26 AM, Mircea Markus wrote: >> >> On Sep 14, 2009, at 12:02 PM, Galder Zamarreno wrote: >> >>> >>> >>> On 09/12/2009 12:31 AM, Michael Neale wrote: >>>> A crazier alternative is to grab the testMethod name and use >>>> that as >>>> part of the key/string - you can do it by getting the stack >>>> trace and >>>> a little bit of work (I had a helper method somewhere). >>>> >>>> Something like: >>>> Thread.currentThread().getStackTrace()(1).getMethodName() (at >>>> the top >>>> of the test method)-- its probably 1 as 0 would be the >>>> getStackTrace >>>> method itself... I have found it helpful when digging through >>>> things >>>> and I see a value from a certain test... Voila! Take your test method and add a java.lang.reflect.Method parameter to it. testng will populate it with the test method information: public void test000(java.lang.reflect.Method method) { System.out.println(method.getName()); } That will give you the test name without doing any fancy stuff :)