[prev in list] [next in list] [prev in thread] [next in thread] 

List:       jacorb-developer
Subject:    Re: [jacorb-developer] meaning of some jacorb error messages
From:       "John I. Seltzer" <john.seltzer () core ! verizon ! com>
Date:       2004-02-24 15:25:28
Message-ID: 403B6CE8.2070200 () core ! verizon ! com
[Download RAW message or body]

Thanks Gerald,

For your info, I've found numerous examples which lead to similar 
confusion because of excess error reporting in the _is_local method, and 
methods that it invokes.

Thanks,
John


Gerald Brose wrote:

>John,
>
>ther is no real error condition in this case (as the functioning of
>your application testifies), the only problem here is that an
>internal JacORB mechanism is abused to check for object existence
>in case of LocateRequests, and this mechanism outputs messages that
>would indicate error conditions in regular operation.
>
>We'll have to tweak the error messages and their creation a little
>here to avoid confusion.
>
>Regards, Gerald.
>
>John I. Seltzer wrote:
>  
>
>>Simon,
>>
>>I don't think the corbaLoc is incorrect, since I know that all processes
>>that use this corbaLoc ARE able to connect to this process.
>>
>>How about this for another theory.  This process regularly tries to sync
>>with it's mate, (same interface, but on a different machine and with a
>>different object key).  I know that jacorb has some code that tries to
>>figure out if an implementation might be local because the interface is
>>the same.  Could it be that whenever this process tries to sync with
>>it's mate, the "is_local" method gets called and tries to find this
>>interface in THIS JVM, and generates the error below.  (I've noted in
>>the past that the _non_existent call will eventually try to connect to
>>the REAL implementation on the other machine, and return return false.
>>
>>Simon McQueen wrote:
>>
>>    
>>
>>>Hi John,
>>>
>>>On 20/02/2004 16:51, John I. Seltzer wrote:
>>>
>>>      
>>>
>>>>Thanks Simon,
>>>>        
>>>>
>>>
>>>You're welcome.
>>>
>>>      
>>>
>>>>This application doesn't use Transient object refs.
>>>>Could it have anything to do with using CorbaLoc to access this service?
>>>>        
>>>>
>>>
>>>If the Corbaloc reference being used is 'wrong' (i.e. has a key for an
>>>object that does not exist in the POA in question), then yes.
>>>
>>>      
>>>
>>>>Simon McQueen wrote:
>>>>
>>>>        
>>>>
>>>>>Hi John,
>>>>>
>>>>>On 19/02/2004 18:04, John I. Seltzer wrote:
>>>>>
>>>>>          
>>>>>
>>>>>>Hi,
>>>>>>
>>>>>>I often see these messages in our log files for one of our
>>>>>>servers.  The server seems to be running properly, and other
>>>>>>processes can communicate with it.  Can anyone shed some light on
>>>>>>what these messages mean?
>>>>>>
>>>>>>
>>>>>>[jacorb.poa] WARN : POA RootPOArid: 167 opname: _non_existent
>>>>>>_invoke: object key not previously generated!
>>>>>>[jacorb] ERROR : org.omg.CORBA.OBJECT_NOT_EXIST: unknown oid
>>>>>>at org.jacorb.orb.giop.ServerRequestListener.deliverRequest(Unknown
>>>>>>Source)
>>>>>>at
>>>>>>org.jacorb.orb.giop.ServerRequestListener.requestReceived(Unknown
>>>>>>Source)
>>>>>>at
>>>>>>org.jacorb.orb.giop.ServerRequestListener.locateRequestReceived(Unknown
>>>>>>Source)
>>>>>>at org.jacorb.orb.giop.GIOPConnection.receiveMessages(Unknown Source)
>>>>>>at org.jacorb.orb.giop.MessageReceptor.doWork(Unknown Source)
>>>>>>at org.jacorb.util.threadpool.ConsumerTie.run(Unknown Source)
>>>>>>at java.lang.Thread.run(Thread.java:512)
>>>>>>
>>>>>>            
>>>>>>
>>>>>Could be something somewhere in your application trying to use stale
>>>>>transient object refs maybe. An OBJECT_NOT_EXIST at that point means
>>>>>that a request has been received with an object key that was not
>>>>>generated by the POA that is supposed to be despatching it.
>>>>>          
>>>>>
>>>>        
>>>>
>>>Cheers,
>>>      
>>>
>
>  
>
>------------------------------------------------------------------------
>
>_______________________________________________
>jacorb-developer maillist  -  jacorb-developer@lists.spline.inf.fu-berlin.de
>http://lists.spline.inf.fu-berlin.de/mailman/listinfo/jacorb-developer
>  
>


[prev in list] [next in list] [prev in thread] [next in thread]