The problem: Trying to load test an application that uses AMF3. Our service replies with session data (and other variables) that must be included in responses to the server.
This means that we cannot pre-compute all of the .binary files and just play them back as they will not have updated session information. I have seen a few responses to this question around the net and wanted to confirm before either giving up or really getting stuck into it. So far we can: Record a session to disk within the JMeter proxy. (.binary files) Replay the session within JMeter and load the files as attachments. View the contents of the AMF replies in human readable format. We have http://code.google.com/p/jmeter-amf-visualizer/ to help decoding the responses – so I’m pretty sure we can extract variables from it. And we have https://github.com/llun/amf or BlazeDS to help encoding to send to the server. So it seems that we have some (if not all of) the pieces, what I am missing is the knowledge of the JMeter architecture. My question is – is it possible to add (preferably in a clean way but I’m not opposed to hacking it in) something between the loading of the binary file from disk to do the variable substitution? And if not is there a good place to add the decode before we write to disk? (Something like what NeoLoad does) Any pointers as to where would be a good place to start would be helpful. Thanks! Alia References: http://mail-archives.apache.org/mod_mbox/jakarta-jmeter-user/200807.mbox/%[email protected]%3e http://mail-archives.apache.org/mod_mbox/jakarta-jmeter-user/200811.mbox/%[email protected]%3e http://mail-archives.apache.org/mod_mbox/jakarta-jmeter-user/200904.mbox/%3c2ab2b3c160bfc14cb0d94d437513c3ba01950...@revmail02.revenuedomain.ie%3e http://mail-archives.apache.org/mod_mbox/jakarta-jmeter-user/201010.mbox/%[email protected]%3e http://note19.com/2008/05/06/how-to-use-jmeter-to-load-test-flex-applications/ http://blog.jteam.nl/2009/07/14/performance-testing-a-flex-blazeds-application/ --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
