[infinispan-dev] JCache implementation documentation
Galder Zamarreño
galder at redhat.com
Mon Apr 29 05:33:35 EDT 2013
On Apr 23, 2013, at 6:03 PM, cotton-ben <ben.cotton at ALUMNI.RUTGERS.EDU> wrote:
>
> Very nice work Galder.
>
> Couple of questions re: Infinispan 5.3.0 ambitions/capabilities to provide
> JSR-107 transactions option.
>
> 1. We noticed on your GitHub branch that you have JCACHE Transactions test
> code. The code uses javax.transaction.TransactionManager (ajuna
> implementation). Will you extend this test to also audition the usage of
> javax.transaction.UserTransaction (arjuna, et. al. implementations?) and its
> JCACHE factory bridge via javax.cache.CacheManager.getUserTransaction() ?
^ Hmmmm, which test are you refering to exactly?
> 2. We notice that your tests include specifying LOCAL isolation
> READ_COMMITTED. Will you extend this to build tests that actually challenge
> Inifispan's integrity for honoring the chosen isolation, i.e. by introducing
> specific "DIRTY_READ" (caching use case) test challenges?
^ TBH, such tests should be added to the TCK. Infinispan already has tests within core/ that verify dirty read situations.
> 3. Can we right now use Infinispan 5.3.0.A to build a test that
> demonstrates Infinispan (via its JTA bridge) honoring a REPEATABLE_READ
> isolation choice?
^ You can build a test anytime :). I don't see why it should not work since we already have REPETEABLE_READ tests in the Infinispan Core testsuite.
> 4. Can you comment on eventually providing tests that demonstrate
> Infinispan 5.3.0 using the TotalOrder (CloudTM?) JTA bridge to achieve
> SERIALIZABLE isolation choice? This one will be especially helpful, as we
> understand Infinispan has the ambition to supporting isolation=SERIALIZABLE
> via the JCACHE API. *We greatly applaud this ambition, and can't wait to
> see it working in Infinispan. * We don't know yet how directly dependent it
> is on TotalOrder being a capability provider (we're studying this).
^ I'm not a total order expert. Pedro, will Total Order enable SERIALIZABLE isolation level? If so, do we already have tests for that?
> 5. Can you post any/all of your JUDcon JSR-107 presentation content?
The presentation will be uploaded to the http://www.jboss.org/events/JUDCon/2013/brazil site shortly, once they've gathered all the presentations. In the mean time, you can find the PDF in https://dl.dropboxusercontent.com/u/6148072/galde-jsr107-brazil.pdf - There's not a huge deal on it since I split the preso to be 50% live coding.
> Again, very nice work ... we'll be following progress here closely. Thanks
^ Thanks for the interest :)
> -Ben
>
>
>
>
> --
> View this message in context: http://infinispan-developer-list.980875.n3.nabble.com/infinispan-dev-JCache-implementation-documentation-tp4026877p4026879.html
> Sent from the Infinispan Developer List mailing list archive at Nabble.com.
> _______________________________________________
> infinispan-dev mailing list
> infinispan-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/infinispan-dev
--
Galder Zamarreño
galder at redhat.com
twitter.com/galderz
Project Lead, Escalante
http://escalante.io
Engineer, Infinispan
http://infinispan.org
More information about the infinispan-dev
mailing list