Hey Mohammed, 2012/6/9 Mohammed Rashad <[email protected]>: > Is there any tool which does simulation testing for Wt apps > I tried siege, ab, JMeter but none are doing simulation properly > > I need to do simulation testing to check the load and max number of users.
I've done this in the past with an online site which uses actual browsers for load testing -- but only tests the initial page load. To test specific actions (setting form data and pushing buttons), this will not be straight forward -- Wt has been designed to actively make this hard because it is a DoS vector. You can make it simpler by given certain widgets (form widgets and buttons) fixed ID's using setId(), but it is still not straight forward to craft the right Ajax requests. > From Browser: > > 1. GET /vmcg?iface=editor&db=vmcg_temp&act=insert&opt=1 > 2. GET /vmcg?wtd=u9YntvaQCNVpACSp&request=style > 3. GET > /vmcg?wtd=u9YntvaQCNVpACSp&sid=487243602&htmlHistory=true&deployPath=%2Fvmcg&request=script&rand=1093612388 > 4. POST /vmcg?wtd=u9YntvaQCNVpACSp > > From JMeter: > > > 1. GET /vmcg?iface=editor&db=vmcg_temp&act=insert&opt=1 > 2. GET /vmcg?wtd=DxvscnwSbMP7TFqI&request=style&js=no > 3. GET /vmcg?wtd=DxvscnwSbMP7TFqI&request=style > > > In both cases Request 1 is same > In JMeter Request 2 has an extra parameter "js=no" and > Request 3 in JMeter is same as Request 2 in Browser > Request 3 &4 in Browser is missing in JMeter. > Also Request 4 is POST > > > My question is there any tool or Can I have the same set of request in both > Browser and JMeter? The question really is -- what do you want to test ? If it is load testing of your Wt application, including the load generated by Wt (rendering), then you will need to simulate actual Ajax sessions. If it is load testing only of backend code (your own code, computations, ...) then you can probably use plain HTML sessions (which are served by default to a JMeter tool which do not execute JavaScript). In the first case, there is a big difference between a new session creation and a simple Ajax update -- usually creating sessions is much more expensive (both CPU and memory-wise) and you can always make sure that your include a heavy operation in the first access if you want to test also load on your backend. Then if your application is online, you are probably best of using a tool like https://browsermob.com/performance-testing or http://loadimpact.com/. In the latter case, you could simply use JMeter or other similar plain HTML tools. They will probably give a slightly pessimistic result since full HTML rendering is usually slower than an Ajax incremental update. Regards, koen ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ witty-interest mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/witty-interest
