Mailing-List: contact users-help@camel.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@camel.apache.org
Received-SPF: pass (nike.apache.org: domain of claus.ibsen@gmail.com
 designates 209.85.223.179 as permitted sender)
MIME-Version: 1.0
In-Reply-To: 
 <b5a176fbfbc34466b013b8b974109258@BLUPR07MB067.namprd07.prod.outlook.com>
References: 
 <4e9ce9c99dca43e29c4cd88457a7ef8b@SN2PR07MB078.namprd07.prod.outlook.com>
 <CAGB5yN=SDwVceJx6fTNMxp8fcaL3wdfGDvAk3BxmNhDGdUKJ-A@mail.gmail.com>
 <b5a176fbfbc34466b013b8b974109258@BLUPR07MB067.namprd07.prod.outlook.com>
From: Claus Ibsen <claus.ibsen@gmail.com>
Date: Tue, 22 Jul 2014 07:32:59 +0200
Message-ID: 
 <CAGB5yNmGiJOKg+mieudQpr8dw20uQENZ0xc5y1eTwQSa3N24gA@mail.gmail.com>
Subject: Re: seda concurrency when sending messages to another route
To: "users@camel.apache.org" <users@camel.apache.org>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Hi

No the direct component does not use the MEP. It just continues
routing the message. And when its done, it returns back to the route
that called it.

eg from A -> to B -> to direct C -> to D

from direct C -> to E -> to F

then when its done at F, then it continues in the 1st route, and goes
to D. And then its done at D, there is no more routes, and the
consumer at A can send back a reply to whoever called it (if its
InOut) or just terminate if its InOnly mode.


On Mon, Jul 21, 2014 at 4:03 PM, Robert Rich <rrich@gsti.net> wrote:
> Thanks Claus!
>
> Is this because it is (as described in the docs) a 'synchronous call to a=
nother endpoint'?   Does the exchange pattern used in the seda route influe=
nce this at all?  If it's InOut, does the Out come from the referenced dire=
ct route?
>
> Sorry for all the questions!
>
> -----Original Message-----
> From: Claus Ibsen [mailto:claus.ibsen@gmail.com]
> Sent: Monday, July 21, 2014 2:18 AM
> To: users@camel.apache.org
> Subject: Re: seda concurrency when sending messages to another route
>
> Hi
>
> Yes you can break up big routes into multiple routes, and "link" them tog=
ether using direct endpoints.
> And yes the seda consumer will not pickup a new message while you call ot=
her routes using direct.
>
> On Sun, Jul 20, 2014 at 6:12 PM, Robert Rich <rrich@gsti.net> wrote:
>> Hi folks,
>>
>> My understanding is that the seda queue is single-threaded by default, p=
rimarily by virtue of 'concurrentConsumers' defaulting to '1'.   To me this=
 means that the route will not pull another message off of the queue until =
the last one has completed processing.  This is my desired behavior.
>>
>> However, the route is getting quite long, and I'm wondering if it's poss=
ible to pass the message to another route while ensuring that the seda cons=
umer will not pull another message off of the queue until that 'invoked' ro=
ute has completed.  I'm assuming this would be done via 'direct', possibly =
with a specific exchange pattern applied to one or both routes?
>>
>> In the case I have in mind, the second/invoked route would end the proce=
ssing chain, but is it possible to have such a pattern in the middle of a r=
oute as well, such that the message is passed to a route and the output of =
that route inserted back into the processing chain of the first route, almo=
st like a subroutine?  It seems like the enricher pattern could be (ab)used=
 this way, but it doesn't seem like that is the intent.
>>
>> Thanks!
>>
>> Bob
>>
>>
>> --
>> Robert Rich
>> CTO/VP
>> Global Security Technologies, Inc.
>> Direct: (614) 291-3456
>> Fax: (614) 356-8078
>> www.gsti.net
>>
>
>
>
> --
> Claus Ibsen
> -----------------
> Red Hat, Inc.
> Email: cibsen@redhat.com
> Twitter: davsclaus
> Blog: http://davsclaus.com
> Author of Camel in Action: http://www.manning.com/ibsen
> hawtio: http://hawt.io/
> fabric8: http://fabric8.io/


--=20
Claus Ibsen
-----------------
Red Hat, Inc.
Email: cibsen@redhat.com
Twitter: davsclaus
Blog: http://davsclaus.com
Author of Camel in Action: http://www.manning.com/ibsen
hawtio: http://hawt.io/
fabric8: http://fabric8.io/