--- sebb <[EMAIL PROTECTED]> wrote: > On 27/10/2007, Richard Hubbell > <[EMAIL PROTECTED]> wrote: > > This is the jmx. So you don't have to go back over > the > > thread on this I was having problems where jmeter > > would send requests to the server like > http://.*/.* > > > > I'd really like to figure out how to have it not > do > > that since it creates a lot of noise in the server > > logs and in the jmeter logs, etc. > > > > If this is hard to read I can re-send as an > > attachment. > > It *is* hard to read - and makes the mail hard for > others to read. > Anyway it is not usable as it stands. > > However, please don't send attachments to the list > either. > > Either store the JMX file on a public server, or > create a Bugzilla > issue and attach the file to that.
I guess bugzilla would be best in this case, I'll do that. In a sense it's unfortunate that the jmx files can't be discussed right here since this is a user group and users seem to have many questions about the jmx since the jmx is everything. Jmeter is the jmx from a user perspective. > > Please make sure that any such test cases are as > small as possible (no > extraneous stuff) and all necessary supporting files > are present. > Also please attach a copy of the jmeter log file > from running the test case. > > In the case of your JMX, I did manage to get the > file to load > (eventually), but it is unusable as there are > several mising files. > > It's not clear that the While Controller is > guaranteed to exit, and > the Link Parser is being applied to the previous > sample in the loop so > it's not surprising that it sometimes does not find > a match - are all > the previous pages guaranteed to contain a link or a > form? It is possible, but what's the alternative for recursing over links on pages? I tried using the feature in the HTTP Request sampler to get only "Embedded URLs must match" regex but that was too limiting. Can you describe in greater detail how complex a regex can go into that field? Would something like this work? href="([^"]+)"|img="([^\s]+)"|imgurl="([^\s]+)" I think you get the idea, there are more than one type of embedded url I'm interested in traversing further. I also tried http://.+/.+ but that didn't work, but I would have thought that would have solved the issue. Maybe it's a missing feature/bug. In regexdom it's a bad idea to use .*, it should be used sparingly. > > Another issue is using 1000 threads with 1 loop just > does not make sense. Does not make sense in general or just to you? It makes sense to me. I would have used 10,000 but the jvm is a bit hungry with memory. There may be some tuning still needed. Stack size, etc. Imagine that each loop does more than one thing. __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
