how to add cpu usage from the tested servers.

[Yuri K]

Hi guys,

is there any plans to extends functionality / charts to add cpu/memory and other metrics from the tested servers?

or any ideas how it is possible to use now ?

thanks,
Yuri

[Chris Price]

That seems like it would be a big change since right now Gatling is a client-side tool.  You'd have to be running some separate Gatling process on the server(s) to track that (though it would be cool).

--
You received this message because you are subscribed to the Google Groups "Gatling User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gatling+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Stéphane Landelle]

This question is very frequently asked on this same mailing list.

Right now, Gatling is only an injector.

You can build a whole monitoring infrastructure with tools such as Nagios, JMXTrans, Logstash, Graphite, etc. and have Gatling intregrate in this system and for example report to Graphite.

Cheers,

Stéphane

2013/5/14 Chris Price <ch...@puppetlabs.com>

[Yuri K]

Understood,

say i would like to integrate monitor data in the gatling charts. is is possible ? What is the best approach for such integration ?

Yuri

Вівторок, 14 травня 2013 р. 18:29:47 UTC+3 користувач Stéphane Landelle написав:

[Stéphane Landelle]

First of all, be aware that you're not allowed to customize Gatling's charting engine for your own needs, because of the Highcharts' licence: https://github.com/excilys/gatling/wiki/License#wiki-highcharts

This said, you need your own charting engine/dashboard.

IMHO, the best approach is to send all your monitoring data to Graphite, and have Gatling do the same with the Graphite integration.https://github.com/excilys/gatling/wiki/Graphite

Regarding how to monitor your system, there's no open-source silver bullet that lets you monitor everything, you'll have to build your monitoring system on top of many components. For example, you can use JMXTrans to monitor JVM's and send to Graphite. This might look like something huge to build, but you should actually be using such tools for monitoring your live platform, and just reuse it for stress testing.

2013/5/15 Yuri K <yko...@gmail.com>