Gigaspaces slow from Magic XPA

Hi,

We use Gigaspaces from Magic XPA as a messaging layer. We have 350 threads which are polling Gigaspaces for what-to-do. It seems that we cannot tune the system further from Magic side, but it still cannot perform more than 8-10 requests/second. I wonder whether it can be tuned further from Gigaspaces side? We have maxThreads = Integer.parseInt( System.getProperty("com.gs.transport_protocol.lrmi.max-threads", "128") ); maxConnPool = Integer.parseInt( System.getProperty("com.gs.transport_protocol.lrmi.max-conn-pool", "1024") ); readSelectorThreads = Integer.parseInt( System.getProperty("com.gs.transport_protocol.lrmi.selector.threads", "4") in services.config.template file.

Should we raise any of these parameters? Or any other? How could we find the bottleneck of the system?

Thank you for any help in advance.

Gyula

asked 2019-04-01 02:55:11 -0500

Gyula gravatar image
edit retag flag offensive close merge delete