On 07/05/2007, Justin C. van Vorst <[EMAIL PROTECTED]> wrote:
>
> What is the correct number of file descriptors necessary to not have ActiveMQ
> fail in a production environment?
It depends on how many connections you're gonna have - I'd set it
pretty high just in case; say to a few thousand.
--
I came across this issue notice :
https://issues.apache.org/activemq/browse/AMQ-1044
https://issues.apache.org/activemq/browse/AMQ-1044
"If async error such as a inactivity IO exception is detected while a
subscription is being created, then that subscription is not fully cleaned
up when the c
What is the correct number of file descriptors necessary to not have ActiveMQ
fail in a production environment?
James.Strachan wrote:
>
> It might be you are just running out of threads; maybe your setting of
> file descriptors per process is too low?
>
>
> On 4/23/07, Justin C. van Vorst <[
Justin C. van Vorst wrote:
> Please don't hijack my thread. Now I'm not sure whether Christopher's
> comment in any way applies to my problem.
If you run out of threads, it could be for a few reasons. You could
have lingering threads in the JVM somewhere (use jstack, kill -QUIT, or
ctrl-\ to get
It might be you are just running out of threads; maybe your setting of
file descriptors per process is too low?
On 4/23/07, Justin C. van Vorst <[EMAIL PROTECTED]> wrote:
On our production server this weekend, activemq-4.1.1 hit a
java.lang.OutOfMemoryError exception and then proceeded to stop
Please don't hijack my thread. Now I'm not sure whether Christopher's
comment in any way applies to my problem.
avin98 wrote:
>
> I have hit an OOMError without using persistent messages, and even if AMQ
> is inactive for a period of about a day or so.
>
> [4/22/07 11:17:44:727 PDT] [WARN]
avin98 wrote:
> I have hit an OOMError without using persistent messages, and even if AMQ is
> inactive for a period of about a day or so.
Did you take a baseline dump after the app reaches a steady state, turn
on heap dumps on memory errors, and then compare the two dumps and see
where the memory
I have hit an OOMError without using persistent messages, and even if AMQ is
inactive for a period of about a day or so.
[4/22/07 11:17:44:727 PDT] [WARN] - [GeronimoConnectionEventListener] -
connectionErrorOccurred called with null
javax.jms.JMSException: Channel was inactive for too long.
