On 8 Jan 2014, at 19:42, uromahn wrote:
> jconsole reports the exact same number, so no issue there.
>
> I re-ran my tests and checking the statistics immediately after the run was
> completed.The statistics make more sense now with the definition you
> provided. The average enqueue time is con
jconsole reports the exact same number, so no issue there.
I re-ran my tests and checking the statistics immediately after the run was
completed.The statistics make more sense now with the definition you
provided. The average enqueue time is consistent with the test - aprox.
1000ms which is the ti
Good point... verify the same numbers you see in hawtio are the same
you see in jconsole.
As for your assumption... the enqueue time is the time spent in the
broker from "when the broker saw the send" to "when the broker
dispatches to a consumer"
On Wed, Jan 8, 2014 at 9:49 AM, James Strachan wro
hawtio should be showing the same values as jconsole (its using the same
underlying JMX MBeans). Do you see the same values in jconsole too?
On 8 January 2014 17:47, uromahn wrote:
> Simple answer: no I haven't had a message sitting around for such a long
> time.
> Also, I must assume that the
Simple answer: no I haven't had a message sitting around for such a long
time.
Also, I must assume that the "enqueue time" is the time it took to put a
message into the queue and not the actual time it spent in a queue, but
please correct me if I am wrong.
Finally, I have a simple test as follows:
the stats are kept for as long as the broker has been up. so for the
life of the broker, would you say you haven't had messages sitting
around on a queue for 15s or greater?
On Tue, Jan 7, 2014 at 3:55 PM, uromahn wrote:
> I've got a simple question with a hopefully simple answer:
>
> the Hawtio
