users@glassfish.java.net

Re: genericra and tibco

From: Marina Vatkina <marina.vatkina_at_oracle.com>
Date: Fri, 16 Jul 2010 16:06:29 -0700

If you have 1 XA resource, it will be committed using (XA) optimized 1PC.

Regards,
-marina

sudharma wrote:
> Hi,
>
> I have configured glassifsh to use Tibco with the help of genericra. I have
> added the property SupportsXA and set it to true. However in the logs I
> still see the following.
>
>
> [#|2010-07-13T14:51:37.781+0530|FINEST|sun-appserver2.1|javax.enterprise.system.core.transaction|_ThreadID=543;_ThreadName=p:
> thread-pool-1; w:
> 13;ClassName=TransactionState;MethodName=setState();_RequestID=68d1f751-f02a-4a4d-9cd8-fda9b27a8709;|Acquiring
> read lock on freeze : state COMMITTING_ONE_PHASE|#]
>
> [#|2010-07-13T14:51:37.781+0530|FINEST|sun-appserver2.1|javax.enterprise.system.core.transaction|_ThreadID=543;_ThreadName=p:
> thread-pool-1; w:
> 13;ClassName=TransactionState;MethodName=setState();_RequestID=68d1f751-f02a-4a4d-9cd8-fda9b27a8709;|Acquired
> read lock on freeze : state COMMITTING_ONE_PHASE|#]
>
> [#|2010-07-13T14:51:37.781+0530|FINEST|sun-appserver2.1|javax.enterprise.system.core.transaction|_ThreadID=543;_ThreadName=p:
> thread-pool-1; w:
> 13;ClassName=RegisteredResources;MethodName=commitOnePhase();_RequestID=68d1f751-f02a-4a4d-9cd8-fda9b27a8709;|Before
> invoking commit_one_phase() on resource:OTSResource : XAResource
> com.sun.genericra.outbound.XAResourceProxy_at_17d2447 XID {XID:
> formatID(4871251), gtrid_length(31), bqual_length(25), data(25000000
> A117FACA 696E6C76 6B35306B 62672C73 65727665 722C5033 33323769 6E6C766B
> 35306B62 672C7365 72766572 2C503333 32372C00)}|#]
>
> [#|2010-07-13T14:51:37.781+0530|FINEST|sun-appserver2.1|com.sun.genericjmsra|_ThreadID=543;_ThreadName=p:
> thread-pool-1; w:
> 13;ClassName=com.sun.genericra.outbound.XAResourceProxy;MethodName=debugxid;_RequestID=68d1f751-f02a-4a4d-9cd8-fda9b27a8709;|Comitting
> outbound transaction with ID
> 696E6C766B35306B62672C7365727665722C50333332372C0025000000A117FACA696E6C766B35306B62672C7365727665722C5033333237|#]
>
> The supports XA should ensure that it does a two phase commit? Is that not
> the case? Any idea as to why this should be happening?
>