chirino_m: you there?

can I use the camel feature in activemq to have broker B consume messages from a destination in broker A and publish it to a destination in broker B?

sure

how do you specify a connection url in camel routes in the activemq.xml?

when I get it going I'll add it to your website

see the example here : http://activemq.apache.org/camel/activemq.html

define as many jms components as you like with different connection factories etc

just give 'em different ids

from("a:someQueue").to("b:anotherQueue")

etc

i got a question about network of brokers: if i have a MDB on machine A listening on queue Q and the same MDB on machine B listening on queue Q, how activemq choose the mdb on A or B to deliver the msg? i mean i know it choose only one, but which is the policy?

and is it possible maybe to define some load balancing policy?

it always chooses local consumers first if they have capacity

jstrachan, so if i'm sending a message from localhost to a queue and i have a consumer deployed on localhost (or attached to the same broker anyways) it has the precedence??

and if all the consumers are remote?

it's random?

it favours consumers on the same broker to remote brokers

jstrachan... does it cost more to publish to a topic if you have more consumers on it?

at publish time not much really - just uses up a bit more cpu/network

network? in what way?

Oh right

sockets

sorry

;)

jstrachan, yeah, i mean, if i have actvemq on AS A networked with Amq on AS b and on AS c, MDB on AS a and AS b, and im sending on the queue from AS c. will Amq deliver the message to one of the mdbs on a or b without any special policy

?

activemq networks means that messages are store/forwarded to available consumers across broker boundaries, yes

Question about ActiveMQ functionality...

I am setting isSendFailIfNoSpace() to true in hopes of a fast failure if DISK is full

However, the code in broker.region.Queue only checks this setting if memoryUsage.isFull() is true

Should I file a bug, or is thiis desired behavior?

FYI, we are configuring a local bridge to consolidate server logs off a large number of application servers. We want to (1) not affect production by filling up disk and (2) not stop threads if allocated disk space is used up

But, we're willing to lose log messages in extreme cases

Anyone here who can answer a question about failover mode

?

I added 1716 and 1717. Please let me know if I didn't take appropriate steps before adding stuff to the JIRA