Re: Websocket + MQTT + NIO: Scaling

[fredri <jf...@pelmorex.com>]

I also tried with Apache Apollo and also get the same results where it seems
there is one thread created per connection.
I would really appreciate if anyone has an idea if I am doing something
wrong, or if there is anything special required to tune ActiveMQ to dispatch
the message asynchronously using a limited number of thread. Or simply
telling me that it is currently by design and expected.
My goal is expose ActiveMQ Websocket directly (or trough a websocket proxy
web server) to the web and I want to achieve maximum scalability and
overcoming the common C10k problem.

Also if you think there is a better place to post by question, let me know.

Thanks in advance,
Freyd 



--
View this message in context: http://activemq.2283324.n4.nabble.com/Websocket-MQTT-NIO-Scaling-tp4681426p4682204.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: Websocket + MQTT + NIO: Scaling

[fredri <jf...@pelmorex.com>]

I have copied the configuration file here:  http://pastebin.com/isJNA8rz
<http://pastebin.com/isJNA8rz>  

I used jconsole to track the number of threads.






--
View this message in context: http://activemq.2283324.n4.nabble.com/Websocket-MQTT-NIO-Scaling-tp4681426p4682282.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.

Re: Websocket + MQTT + NIO: Scaling

[surfnerd <su...@outlook.com>]

Can you please upload your full configuration file for AMQ ?
I never tried performance on Java 8 though.



--
View this message in context: http://activemq.2283324.n4.nabble.com/Websocket-MQTT-NIO-Scaling-tp4681426p4682275.html
Sent from the ActiveMQ - User mailing list archive at Nabble.com.