Return-Path: Delivered-To: apmail-httpd-test-dev-archive@httpd.apache.org Received: (qmail 20145 invoked by uid 500); 24 Sep 2002 09:35:54 -0000 Mailing-List: contact test-dev-help@httpd.apache.org; run by ezmlm Precedence: bulk Reply-To: test-dev@httpd.apache.org list-help: list-unsubscribe: list-post: Delivered-To: mailing list test-dev@httpd.apache.org Received: (qmail 20121 invoked from network); 24 Sep 2002 09:35:49 -0000 Date: Tue, 24 Sep 2002 11:30:38 +0200 From: Jacek Prucia To: test-dev@httpd.apache.org Subject: Re: flood errors with login ASP script Message-Id: <20020924113038.09f69f30.jacek.prucia@7bulls.com> In-Reply-To: References: Organization: 7bulls.com S.A. X-Mailer: Sylpheed version 0.8.2 (GTK+ 1.2.10; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N On Mon, 23 Sep 2002 11:52:58 -0700 (PDT) Erik Curiel wrote: Important thing first: > We're running the latest non-CVS version of flood, compiled with SSL > support, on Solaris 8, SPARC Ultra5. I assume, you're using flood-0.4 tarball from ASF distribution site. We have checked in a lot of bugfixes/features since that release. You schould checkout current CVS version as described on flood site, http://httpd.apache.org/test/flood, section CVS access. > Hi! I'm a flood newbie, trying to use flood for some very basic > load-testing on a new commercial site we're building. We want to use > the url-list feature to mimic actual click-through paths that we think > clients will use on the site. Clients must log in to use the site, > however. We can hit the first page, the login form, with flood with > no problem. When a client enters his or her login info > (username/password), the form posts to an ASP script that verifies the > info, sets a cookie with the session ID and then redirects (via the > ASP Session.Redirect function, i.e. a Refresh HTTP header is > generated) to the site's home page. When we try to post directly to > this ASP script with login info, flood reports a failure, but gives no > info on what exactly failed. You schould manually check headers returned by server (curl/netcat/telnet). Flood's verify_200 function (which you are using) expects HTTP/1.1 response. If (for some reason) your server returns HTTP/1.0, you can use other verify function -- verify_status_code -- like this: verify_status_code Again -- you'll have this functionality only if you check out latest flood source code. > I try to write output to a > log-file, by specifying a logfile element in the XML config file, but > no log file appears to be created (at least, I can't find it). Yup. I think that logfile element was a part of design, but never really got implemented. However I was thinking about such functionality myself. I'll put it on my flood TODO list. > When we look > at the log file on the server we're trying to hit, it shows that flood > did indeed post to it, and the server sent back a 302 ("Object Moved) > message, with the location of the page that the client is being > redirected to. When I use the connection "keep-alive" directive in the > header, by declaring the appropriate XML tag in the config file, then > flood hits the login form no problem, but then hangs when it tries to > hit the login script. The server log still shows, however, that flood > did succeed in posting to the script and the server responded with the > same 302. A small note here: after reading your XML file i see, that login POST is the last url in your urllist. Using hints above you can get flood to report 'OK', but it will *not* go further. I mean: it will not hit the url received with 302 response. You have to add that url to your XML file by hand. In near future, flood will have a 'mimic_browser' switch, which when turned on (among other things) will cause flood to respond to 3xx type responses. However right now, you'll have to do this by yourself. > I insert the XML file we're using at the end of this email. (We're > not really trying to hit test.com, of course, but we're building this > site for a big bank so we can't give out the real URL.) Umm... There's another new feature in flood, that you might find usefull. You can have following config: http://www.your_bank.com /foo.asp /bar.asp With a little change in baseurl tag you can test different instances of your site (like: test.your_bank.com, debug.your_bank.com and so on) with the same url list. As you can see -- flood has a lot of nice features, but we need a real user documentation. This is exactly what I'm working on right now, so stay tuned. regards, -- Jacek Prucia 7bulls.com S.A. http://www.7bulls.com/