Oh!  I forgot, how about client-side, you use a database of some sort
(this[1] is how it would look service-side, but the logic is quite similar) 
that records a true/false or whatever that the transaction finished
successfully.  Then upon rebooting, your client can read that database table
to determine which transactions need to be resent.

HTH,
Glen

[1] http://www.jroller.com/gmazza/date/20080308


Hyacinthe Meniet wrote:
> 
> Hi Glen,
> 
> From what I understand, there are 2 options to WS-RM, persistent and non
> persistent:
>  
> If I use non-persistent, it does not solve my problem since the client
> will loose its RM context during the reboot
> 
> If I use persitent, it requires the same MOM on both ends of the wire, and
> this is an unacceptable constraint; let's say we want to communicate
> between CXF on e.g. WebSphere and WCF on .NET. WebSphere would implement
> WS-RM persistency with its embedded SIB and WCF with MSMQ. Deadlock!
> So we need some client toolkit only solution.
> 
> 
> 
> 

-- 
View this message in context: 
http://www.nabble.com/How-to-persist-and-restore-a-cxf-client-context-tp19368970p19414256.html
Sent from the cxf-user mailing list archive at Nabble.com.

Reply via email to