Nenufar,
i am no expert here...
but i view it this way:
- scripts are for simulating something that the user might do
- could be simply hitting the "about" page if you want to
simulate DOS attack :-)
- could be logging in, entering data, searching for data, etc
- threads are for simulating number of users
- simulating multiple simultaneous users
- simulating different users groups to run different scripts
you can try different approaches to see the effect.
then choose what works best.
for example... you could try
- [user_group-1]
threads: 10
script: AW_prodtest.py
and then try
- [user_group-1]
threads: 10
script: about_test.py
- [user_group-2]
threads: 10
script: login_and_search_test.py
and see how things look...
then bump up the number of threads and see the effect
you will likely see it reach a point where the throughput
(requests/second) drops off and the data gets "noisy" which could
mean the system is getting unstable.
i view multi-mechanize as a tool to help see how the system
responds. and if it does not respond as you expect, you have some
work to do to explain why :-)
hope this helps
jon
blog: http://technicaldebt.com
twitter: http://twitter.com/JonKernPA
Nenufar said the following on 1/23/12 6:13 AM: