[omniNotify] Consumer failure handling?

Robert E. Gruber gruber at research.att.com
Fri Dec 19 12:56:53 GMT 2003


Cameron, you are exactly right, a communication failure causes the channel to 'give up' on
a consumer.  There are two modes for connection reliability, 'best effort' and
'persistent', where omniNotify does not support the persistent option.  It is possible
that the current implementation of best effort delivery gives up too soon... I could look
at making it try a certain number of additional times before 'giving up.'  

However, I will be traveling until January 2nd, so I will need to wait until then to look
at it.

Happy holidays everyone!
-- Bob


-----Original Message-----
From: omninotify-list-bounces at omniorb-support.com
[mailto:omninotify-list-bounces at omniorb-support.com] On Behalf Of Cameron Rochester
Sent: Thursday, December 18, 2003 8:06 PM
To: omninotify-list at omniorb-support.com
Subject: [omniNotify] Consumer failure handling?

Hi all,

If the NotificationService fails to push an event to a
StructuredPushConsumer for any reason (say a TRANSIENT) will it
automatically assume that the consumer is no longer valid and clean up? Or
will it retry the event a couple of times before giving up and flagging the
consumer offline?

I ask this because we are seeing consumers occassionally stop receiving new
events. This behaviour is obvious as we run multiple instances of an
application, on the same workstation, that all subscribe identically.
Occasionally we will observe that one of these applications will stop
receiving new events while the others will continue quite happily.

Very strange behaviour, and very difficult to debug.

Thanks
Cameron

_______________________________________________
omninotify-list mailing list
omninotify-list at omniorb-support.com
http://www.omniorb-support.com/mailman/listinfo/omninotify-list




More information about the omninotify-list mailing list