You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@servicemix.apache.org by sachin2008 <es...@gmail.com> on 2008/03/05 06:26:47 UTC
Re: Reg:Fault Handling
Hi gnodet,
I have a doubt in aggregation of messages using split aggregator.
Use case which i have explained previously:
jms consumer --->static receipient list----->(3 lw-container
components)--------> split aggregator----> jms provider
I need some clarification on the aggregation of messages using split
aggregator.
1.whenever any fault happens in the one of the 3 lw-container components
then how split aggregator will aggregate the messages.
2. since we are assigning properities like correalation id,count and index
so whenever there is any fault happens in one of the component then split
aggregator throws an exception like: unable to
retrieve correlation id.
can anyone please help me in this regard....
gnodet wrote:
>
> If tehre is a fault in echo1 component, the routing slip will be aborted
> and
> the fault will be sent back to the consumer, so echo2 won't receive it.
> A possible way would be to use the faultsTarget of the EIP pipeline to
> redirect faults somewhere else for a different processing.
> In any case, the aggregator must either receive faults, recognize them
> and process them as needed, or should timeout when all the messages
> are not received and send a fault itself. The first solution would be
> better.
>
> On Thu, Feb 28, 2008 at 9:37 AM, sachin2008 <es...@gmail.com> wrote:
>>
>> Hi,
>>
>> How to handle faults in case of EIP component..
>>
>> Let's suppose if we will teke static routing slip.
>>
>> <eip:static-routing-slip service="replaceMe:routingSlip"
>> endpoint="endpoint">
>> <eip:targets>
>> <eip:exchange-target service="replaceMe:echo1" />
>> <eip:exchange-target service="replaceMe:echo2" />
>> </eip:targets>
>> </eip:static-routing-slip>
>>
>> In this routing slip will forward an inout message to replaceME:echo1
>> and
>> then from replaceME:echo1 message will route to replaceMe:echo2.
>> If there is any fault message will return from replaceMe:echo1 to
>> replaceMe:echo2 then how we can handle these type of faults.
>>
>>
>> Now let's consider we are using aggregator for aggregate the messages
>> coming
>> from different components. If there is any fault occured in any of the
>> component then aggregator how it handles the fault before aggregating
>> all
>> the messages.
>>
>>
>> Can anyone please put light on this fault handling
>>
>> -----
>> Cheers
>> Praveen Oruganti
>> "Think before you act and act on what you believe"
>> --
>> View this message in context:
>> http://www.nabble.com/Reg%3AFault-Handling-tp15731923s12049p15731923.html
>> Sent from the ServiceMix - User mailing list archive at Nabble.com.
>>
>>
>
>
>
> --
> Cheers,
> Guillaume Nodet
> ------------------------
> Blog: http://gnodet.blogspot.com/
>
>
-----
Cheers
Praveen Oruganti
"Think before you act and act on what you believe"
--
View this message in context: http://www.nabble.com/Reg%3AFault-Handling-tp15731923s12049p15843534.html
Sent from the ServiceMix - User mailing list archive at Nabble.com.
Re: Reg:Fault Handling
Posted by Guillaume Nodet <gn...@gmail.com>.
Faults are usually used with InOut exchanges and not with InOnly.
If any of your lightweight component thrown an exception, it should be
set on the exchange with an ERROR status and the exchange sent back
to the static recipient list. I don't really follow how the aggregator would
receive faults.
On Wed, Mar 5, 2008 at 6:26 AM, sachin2008 <es...@gmail.com> wrote:
>
> Hi gnodet,
> I have a doubt in aggregation of messages using split aggregator.
>
> Use case which i have explained previously:
>
> jms consumer --->static receipient list----->(3 lw-container
> components)--------> split aggregator----> jms provider
>
> I need some clarification on the aggregation of messages using split
> aggregator.
> 1.whenever any fault happens in the one of the 3 lw-container components
> then how split aggregator will aggregate the messages.
> 2. since we are assigning properities like correalation id,count and index
> so whenever there is any fault happens in one of the component then split
> aggregator throws an exception like: unable to
> retrieve correlation id.
>
> can anyone please help me in this regard....
>
>
>
>
>
> gnodet wrote:
> >
> > If tehre is a fault in echo1 component, the routing slip will be aborted
> > and
> > the fault will be sent back to the consumer, so echo2 won't receive it.
> > A possible way would be to use the faultsTarget of the EIP pipeline to
> > redirect faults somewhere else for a different processing.
> > In any case, the aggregator must either receive faults, recognize them
> > and process them as needed, or should timeout when all the messages
> > are not received and send a fault itself. The first solution would be
> > better.
> >
> > On Thu, Feb 28, 2008 at 9:37 AM, sachin2008 <es...@gmail.com> wrote:
> >>
> >> Hi,
> >>
> >> How to handle faults in case of EIP component..
> >>
> >> Let's suppose if we will teke static routing slip.
> >>
> >> <eip:static-routing-slip service="replaceMe:routingSlip"
> >> endpoint="endpoint">
> >> <eip:targets>
> >> <eip:exchange-target service="replaceMe:echo1" />
> >> <eip:exchange-target service="replaceMe:echo2" />
> >> </eip:targets>
> >> </eip:static-routing-slip>
> >>
> >> In this routing slip will forward an inout message to replaceME:echo1
> >> and
> >> then from replaceME:echo1 message will route to replaceMe:echo2.
> >> If there is any fault message will return from replaceMe:echo1 to
> >> replaceMe:echo2 then how we can handle these type of faults.
> >>
> >>
> >> Now let's consider we are using aggregator for aggregate the messages
> >> coming
> >> from different components. If there is any fault occured in any of the
> >> component then aggregator how it handles the fault before aggregating
> >> all
> >> the messages.
> >>
> >>
> >> Can anyone please put light on this fault handling
> >>
> >> -----
> >> Cheers
> >> Praveen Oruganti
> >> "Think before you act and act on what you believe"
> >> --
> >> View this message in context:
> >> http://www.nabble.com/Reg%3AFault-Handling-tp15731923s12049p15731923.html
> >> Sent from the ServiceMix - User mailing list archive at Nabble.com.
> >>
> >>
> >
> >
> >
> > --
> > Cheers,
> > Guillaume Nodet
> > ------------------------
> > Blog: http://gnodet.blogspot.com/
> >
> >
>
>
> -----
> Cheers
> Praveen Oruganti
> "Think before you act and act on what you believe"
> --
> View this message in context: http://www.nabble.com/Reg%3AFault-Handling-tp15731923s12049p15843534.html
>
>
> Sent from the ServiceMix - User mailing list archive at Nabble.com.
>
>
--
Cheers,
Guillaume Nodet
------------------------
Blog: http://gnodet.blogspot.com/