Yes, I didn't realize that remove returns an object I asumed that it returns true or false. Congrats for the optimization. Thanks.
Cheers Zacarias Monzon. On 10/26/07, Guillaume Nodet <[EMAIL PROTECTED]> wrote: > > Why not using: > > exchange = exchanges.remove(exchange.getExchangeId()); > > instead of > > exchange = exchanges.get(exchange.getExchangeId()); > exchanges.remove(exchange.getExchangeId()); > > It should be the same but in an atomic call. > > On 10/26/07, Zacarias <[EMAIL PROTECTED]> wrote: > > It seems that patch files are filtered or something. I attached the > patch > > file at https://issues.apache.org/activemq/browse/SM-822 > > > > Cheer, > > Zacarias > > > > On 10/26/07, Guillaume Nodet <[EMAIL PROTECTED]> wrote: > > > > > > Could you raise a JIRA and attach your patch ? It seems to have been > > > removed from your email. Another solution is to use the nabble forums > > > too. > > > > > > On 10/26/07, Zacarias <[EMAIL PROTECTED]> wrote: > > > > I implemented a posible solution for the "component not owner" > problem. > > > I > > > > found that > > > > > > > > deployables/bindingcomponents/servicemix-http/src/main/java/org/apache/servicemix/http/processors/ConsumerProcessor.java > > > > sleeps meanwhile the jsr componet modifies the message.In JMSFlow > the > > > > variable "exchange" is serialized to an MQ queue and deserialized > that's > > > why > > > > when ConsumerProccesor wakes up exchange is in the wrong state. With > > > this > > > > patch I propose to take again the "exchange" from the "exchanges" > list. > > > I > > > > tested it with Seda flow and JmsFlow separated and it works. > > > > I'm looking forward for your answer. > > > > > > > > Best regards. > > > > > > > > Zacarías Monzón > > > > > > > > > > > > > > > > > -- > > > Cheers, > > > Guillaume Nodet > > > ------------------------ > > > Blog: http://gnodet.blogspot.com/ > > > > > > > > -- > Cheers, > Guillaume Nodet > ------------------------ > Blog: http://gnodet.blogspot.com/ >
