On Tue, 01 Mar 2005 12:58:04 -0600, Danny R. Faught <[EMAIL PROTECTED]> wrote: > sebb wrote: > > Normally a single thread represents a single user, so this would mean > > first time for each thread (user) in a Thread Group within a test run. > > This fits in with the way the Cookie Manager works. > ... > > It might be a lot harder to implement a reset control, and I'm not > > sure what it would correspond to in terms of a test. > > I'm trying to learn what represents a distinct virtual user and a single > session on a site. From what you said in the first paragraph, it looks > like a thread represents a user, and represents the same user on > repeated loops in the thread group.
Yes, that's the way it is normally used - testing with independent users requires independent threads. However, one does not have to stick with the same user for all the loops of the thread. One can set up a test to login at the start, and logout at the end of the script. This would allow for a test with more users than threads. In which case one would not need the "first-time" switch - as far as I can tell. > A reset would facilitate reusing a thread to act as a brand new user, > but now I'm thinking that that goes against the design of the tool. Yes. > >>Mostly, the tool just needs to track which items have been downloaded > >>already - it doesn't need to cache the contents, except that some > >>listeners need to show them. I guess they're already caching things but > >>not reusing them. > > > > Not sure I understand the last sentence - who is they? > > The listeners that record the content sent back from each request. The results are passed to each of the Listeners in turn and then discarded - they are not cached in the normal sense of a cache. > > A proper cache ought to take note of the relevant Header lines and > > expiry dates etc. > > Yeah, I shouldn't assume that caching is simple. > -- > Danny R. Faught > Tejas Software Consulting > http://tejasconsulting.com/ > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
