RE: CorrelationId

xavier wrote Perhaps I must free some object, but in debug mode and checking the memory, memory grow up, each time I do pn_messenger_subscribe, pn_messenger_recv, etc. did you try running your binary via valgrind memcheck? -- View this message in context:

RE: CorrelationId

Hi, Any body have an idea about this? The new version of the lib will fix this, or it's my code? regards -- View this message in context: http://qpid.2158936.n2.nabble.com/CorrelationId-tp7614606p7617531.html Sent from the Apache Qpid Proton mailing list archive at Nabble.com.

Re: CorrelationId

Hi Dominic, I set PN_TRACE_FRM=1 and here my trace. [006FC110]: - SASL [006FC110]:0 - @sasl-init(65) [mechanism=:PLAIN, initial-response=b\x00emc2\x00emc2] [006FC110]: - SASL [006FC110]:0 - @sasl-mechanisms(64) [sasl-server-mechanisms=@PN_SYMBOL[:ANONYMOUS, :PLAIN]] [006FC110]:0 -

Re: CorrelationId

Hi Xavier, so 006FE360 is your messenger that is responsible for subscribing and receiving the messages based upon the selector. xavier wrote [006FE360]: - SASL [006FE360]:0 - @sasl-init(65) [mechanism=:PLAIN, initial-response=b\x00emc2\x00emc2] [006FE360]: - SASL [006FE360]:0 -

RE: CorrelationId

Here the right code //Doing this one time pn_messenger_t* msgConsumer= pn_messenger(NULL); pn_messenger_set_timeout (msgConsumer, 1000); pn_messenger_set_blocking (msgConsumer, true); pn_messenger_set_incoming_window (msgConsumer, 1); // Doing this more time pn_link_t* link;

RE: CorrelationId

, Xavier Subject: Re: CorrelationId Hi Xavier, so 006FE360 is your messenger that is responsible for subscribing and receiving the messages based upon the selector. xavier wrote [006FE360]: - SASL [006FE360]:0 - @sasl-init(65) [mechanism=:PLAIN, initial-response=b\x00emc2\x00emc2] [006FE360

Re: CorrelationId

Hi Xavier, xavier wrote I posted my solution, it works, but I did some performance test, and it's not very good, I found why. In my solution, I start, and stop for any receiver the messenger, because if I don't do this, with my code (posted) I receive one time, the message with a filter

Re: CorrelationId

Hi Dominic unfortunately, it 's does not work!! I try your idea, see my code (I move the pn_link_detach() and pn_link_open() in all my code, but .) Here my code: // Init one time to use many time pn_messenger_t* msgConsumer= pn_messenger(NULL); pn_messenger_set_timeout (msgConsumer, 1000);

Re: CorrelationId

Hi all, I found the solution (for ActiveMq, with a connector in AMQP, and a connector with openwire), so I share the solution. I used the idea of Dominic Evans, so If I want to get a message from a filter (for example correlationId ), I did this: pn_messenger_t* msgConsumer= pn_messenger(NULL);

Re: CorrelationId

Hi all, I re start to work on this issue!!! and it is not work! So here my last code: // get recv link pn_link_t* link = pn_messenger_get_link (messengerConsumer, myAdress, false); // get the source link, is't it, I saw, before doing that on the the activemq's log, than the filter is: the

Re: CorrelationId

Hi all, pn_messenger_get_link is not in the 0.7 qpid lib, right? Is it in the next release 0.8 ? If yes do you have any date for it, and code sample, with? Thanks a lot regards -- View this message in context: http://qpid.2158936.n2.nabble.com/CorrelationId-tp7614606p7614826.html Sent

Re: CorrelationId

fadams wrote On 08/10/14 18:49, Rafael Schloming wrote: Actually I think it is possible to do this with messenger now that dominic's patches are in. You can access the link associated with any given address using pn_messenger_get_link() and configure it however you want to. --Rafael Oooh

Re: CorrelationId

On Thu, Oct 9, 2014 at 8:44 AM, Dominic Evans dominic.ev...@uk.ibm.com wrote: fadams wrote On 08/10/14 18:49, Rafael Schloming wrote: Actually I think it is possible to do this with messenger now that dominic's patches are in. You can access the link associated with any given address

Re: CorrelationId

Hi Frase, Thanks for your explanation, but here, my code: In the requester: char * corrId; . . pn_bytes_t bytes = pn_bytes(correlationId.size(), corrId); pn_atom_t id; id.type = PN_STRING; id.u.as_bytes = bytes; pn_message_set_correlation_id(message, id); and after, send it

Re: CorrelationId

On 08/10/14 08:16, xavier wrote: Hi Frase, Thanks for your explanation, but here, my code: In the requester: char * corrId; . . pn_bytes_t bytes = pn_bytes(correlationId.size(), corrId); pn_atom_t id; id.type = PN_STRING; id.u.as_bytes = bytes; pn_message_set_correlation_id(message,

Re: CorrelationId

On Wed, Oct 8, 2014 at 1:30 PM, Fraser Adams fraser.ad...@blueyonder.co.uk wrote: On 08/10/14 08:16, xavier wrote: Hi Frase, Thanks for your explanation, but here, my code: In the requester: char * corrId; . . pn_bytes_t bytes = pn_bytes(correlationId.size(), corrId);

Re: CorrelationId

On 08/10/14 18:49, Rafael Schloming wrote: On Wed, Oct 8, 2014 at 1:30 PM, Fraser Adams fraser.ad...@blueyonder.co.uk wrote: On 08/10/14 08:16, xavier wrote: Hi Frase, Thanks for your explanation, but here, my code: In the requester: char * corrId; . . pn_bytes_t bytes =

Re: CorrelationId

Hi all, Thanks you for your reply, help? So if I understand properly, to do that with messenger is not trivial??? But I don't understand why, because with the pattern request/reply in asynchronous mode, and with dedicated queue, we must have this feature. isn't it? So to do that must I take some

Re: CorrelationId

On 07/10/14 08:04, xavier wrote: Hi all, Thanks you for your reply, help? So if I understand properly, to do that with messenger is not trivial??? But I don't understand why, because with the pattern request/reply in asynchronous mode, and with dedicated queue, we must have this feature. isn't

Re: CorrelationId

ActiveMQ currently uses an AMQP filter to specify the selector string the broker should use for the consumer. I dont believe Messenger currently supports specifying filters, however I have next to no experience with Messenger so perhaps someone with more knowledge can comment. Robbie On 2