Yes, I suppose that's another option, or wrap things with a
DelegateProcessor.
Our route builder (on top of the Camel RouteBuilder) is part of a larger
framework and makes all sorts of dynamic and configured decisions how to
assemble various scenarios.
Cheers!
--
View this message in context:
h
Ok, so I initially used #1 but it failed to build the route. The issue seems
to be different than what I originally assumed...
Basically I am building a route and want to insert a processor based upon a
condition (which is: does the processor exist?). Initially I did this with
an if-then-else stat
[Camel 2.7.2]
Sorry, but I am not quite clear about the syntax of inserting Predicates,
esp to dynamically wire in processors... I am also not sure about the
significance of end() and the newer endChoice(). Basically, doing 1 choice
seems to be working but 2, as in below examples, breaks the conti
Alright, I coded a simple solution basically using a bounded LinkedList as an
LRU cache for filtering already processed message IDs on the callback.
This is not the most efficient approach (overhead piping payload through
Spring/Camel, before reaching my filtering logic & potential for contention
Thanks Claus -
In the meantime, do you have an opinion how to work around it? I see two
options (a) go via concurrentConsumers and somehow filter duplicate messages
or (b) hand the payload off to a custom async worker thread pool. (a) seems
to be quicker to do but not so sure about the overhead o
Hi Tim -
Using Camel 2.7.2.
I actually had the ExchangePattern before threads() originally - the above
is just one permutation of several I tried (just moved it back - same
result).
Yes, I guess it should block if the 'consumer'/publisher is expecting a
response. But it's using OutOnly ExchangePa
I sometimes use AMQ for isolated local testing but our production transport
is Tibco EMS.
But in any case, the Camel approach to multi-threading doesn't seem to be
working as expected, so I am not quite ready to make changes to the
underlying configuration. I could probably also set up a single li
Hey Tim.
Sorry, I have everything parameterized and missed a few values. The
ExchangePattern is InOnly (system exposes same behavior for OutOnly for what
it's worth...). sleepTime is constant 1000 in this case - you see nicely in
the log how threads have a 1sec gap between ''start' and 'end' log a
Right, if I set concurrentConsumers=1 and while building the route
threads(10) I do end up with a thread pool, however, these threads are used
serially.
The configuration is as follows:
final String from =
"jms:topic:esb_test_topic?concurrentConsumers=1";
final String to = "bean
Ok, makes sense. Cheers!
--
View this message in context:
http://camel.465427.n5.nabble.com/Each-concurrentConsumer-on-JMS-Topic-receives-ALL-msgs-tp4558687p4558845.html
Sent from the Camel - Users mailing list archive at Nabble.com.
Hi -
The answer may be as simple as 'that's the way it is, but what we observe is
not ideal:
Basically a pool of 'listeners'
(jms:topic:super_topic?concurrentConsumers=10) all receive the same message,
whereas the ideal scenario would be only one thread handles each msg and to
process different m
Yes, I am dumbing down the scenario and don't mention recovery, threading,
etc to focus on the concrete issues how to link up a temp channel. The
client specified a requirement of one defined queue per service - I
currently have a workaround with one queue and a topic but it's not fully
meeting the
Brackets were removed from the example in the previous post:
(jms:temp:queue:[mytempqueuename])
--
View this message in context:
http://camel.465427.n5.nabble.com/Issue-with-JMS-Temp-Queues-Tibco-EMS-tp3425694p3779620.html
Sent from the Camel - Users mailing list archive at Nabble.com.
Thanks Christian -
In a nutshell and simplified view, ignoring threads etc, what we need is
similar to request/response but keeping the temp response channel open as
long as we wish, to send several messages. In fact I don't understand the
purpose of a Camel Temp queue if different parties don't c
Hallo Christian -
Right, we are using temp queues for responses, but need more control than
standard req/resp. The response may not just be synchronous but result in
several messages (this is not uncommon in the financial/analytics world and
in general helps creating dynamic interfaces that incre
Hi, we don't seem to be receiving messages on temp queues. The underlying
transport is Tibco EMS (JMS).
Below are snippets from the log regarding listener and publisher starting
up...
It looks all fine (to my eye..) besides the 'real' temp queue only being
mentioned on the sender side
(Queue[$TMP$.
Ok, so I'll use:
029public void process(Exchange exchange) throws Exception {
030// mark the exchange to stop continue routing
031exchange.setProperty(Exchange.ROUTE_STOP, Boolean.TRUE);
032}
Thanks for the quick response! :)
--
View this message in contex
We are extending DelegateProcessor to fiddle with the payload before it hits
the business logic.
How do we correctly stop an Exchange that we want to drop?
Would it be this or is there a more straight forward method?
myExchange.getContext().getInFlightRepository().remove(myExchange)
Thanks!
-
Thanks, that seems to help!
I did actually read the docs while investigating the issue but it wasn't too
clear:
http://camel.apache.org/jms.html
disableReplyTo falseIf true, a producer will behave like a InOnly exchange
with the exception that JMSReplyTo header is sent out and not be suppre
Nope, too early. It's still creating temp reply listeners on OutOnly routes
2011-02-04 01:10:00,784 | INFO | ent.jms.reply.TemporaryQueueReplyManager |
doStart | ead #2 - Threads | Starting reply listener container on
endpoint: Endpoint[jms://topic:esb_topic_status]
2011-02-04 01:10:00
Right, looks like this was a bug and fixed in 2.6. Cheers.
--
View this message in context:
http://camel.465427.n5.nabble.com/After-publish-OutOnly-Reply-received-for-unknown-correlationID-tp3368679p3370239.html
Sent from the Camel - Users mailing list archive at Nabble.com.
Hi.
We are sending a 'heart beat' on a JMS topic with am OutOnly publisher. This
results in local log WARNs for related messages. Are we configuring this
incorrectly or how can we avoid these problems?
Many thanks. / m
The route is defined/initialized as
from(from).setExchangeP
>
from(source).process(myInProcessor).to(destination).process(myOutProcessor).
Ah, this works. Very nice. Thanks!
--
View this message in context:
http://camel.465427.n5.nabble.com/JAXB-causing-Unit-Test-Failure-without-any-actual-marshalling-active-tp3278807p3286117.html
Sent from the Camel -
Willem - I should upgrade to 2.5 first and see if I can reproduce but it's
not as straight forward as hoped I am afraid.
Btw - I am defining the routes programmatically at startup time such as -
...
RouteBuilder {
public void configure() {
from(source).process(myProcessor).to(destination);
Oh, I wasn't aware 2.5 is out. We are on 2.4 still.
I don't think I'll get to a test project any time soon as I'm currently
working through holidays to meet deadlines (although having been informed it
is considered treason to work on Thanksgiving in the US).
The workaround and my preference is a
@Willem
The payload is a simple JAXB annotated Pojo (In this case a Status message,
containing some strings and maps).
To recap:
1. Unit tests (and actual operation) run fine without maven dependency of
camel-jaxb.
2. Jaxb itself runs fine on the classes. Schema is generated and unit tests
pass
Thanks for following up. In the meantime I have moved on and created a custom
Processor chained into the route, which provides more flexibility for other
processing.
The rest of the stack trace just went further into the jaxb RI and didn't
seem to provide additional clues why Camel is picking thi
Hi.
We are planning to plug JAXB marshalling into our Camel routes.
To start with I annotated DTOs for JAXB and create a schema for .NET
clients. Next was to get Camel involved.
However, simply adding the Maven dependency is causing the unit tests to
throw execptions as below.
Again - marshal/u
28 matches
Mail list logo
