2009/2/16 Rafal Rusin <[email protected]>:
> 2009/2/13 Matthieu Riou <[email protected]>:
>> On Fri, Feb 13, 2009 at 2:10 AM, Rafal Rusin <[email protected]> wrote:
>>
>>> Hello,
>>>
>>> could you explain me a bit bpel execution regarding BpelRuntimeContext?
>>> I saw a following happened:
>>>
>>> 09:59:25,185 | INFO  | pool-4-thread-4 | BpelRuntimeContextImpl   |
>>> .engine.BpelRuntimeContextImpl  160 | BpelRuntimeContextImpl created
>>> for instance 20598 24589427
>>> 09:59:25,269 | INFO  | pool-4-thread-7 | BpelRuntimeContextImpl   |
>>> .engine.BpelRuntimeContextImpl  160 | BpelRuntimeContextImpl created
>>> for instance 20598 1821816
>>> 09:59:25,628 | INFO  | pool-4-thread-4 | BpelRuntimeContextImpl   |
>>> .engine.BpelRuntimeContextImpl  843 | BpelRuntimeContextImpl setting
>>> execution state on instance 20598 24589427
>>> 09:59:25,781 | INFO  | pool-4-thread-7 | BpelRuntimeContextImpl   |
>>> .engine.BpelRuntimeContextImpl  843 | BpelRuntimeContextImpl setting
>>> execution state on instance 20598 1821816
>>>
>>
>> Is that on trunk or 1.x ? In any case I'm puzzled as to how this is
>> possible. There's a big lock on instances so that a given instance can't be
>> executed by two threads in parallel. Check BpelEngineImpl.onScheduledJob.
>> Given that an INVOKE_RESPONSE handling goes through this method, it's
>> surprising. Logging on the lock manager could provide more details.
>
> OK, thanks for clarifying, I'll check it and give more info.
>

This was on 1.x, in JBI.

Rafal

Reply via email to