hi Mike, a nice way to debug this would be to print out the data being sent back and forth. If you want, you can put my java loadbalancer in front, and set it to debugString and it will print out the HTTP packages being sent.
http://www.filip.net/tomcat-clustering.html at the end of the page, there is a loadbalancer, there is a flag "-debugString" that will print out every request and response to standard out. That way you can see what your cookies are set too. can you give us an example of how your cluster hostname is different from a hostname on any other machine, and also, are you running Apache in front of Tomcat? Filip > -----Original Message----- > From: hegerich, mike [mailto:[EMAIL PROTECTED] > Sent: Wednesday, March 19, 2003 9:50 AM > To: 'Tomcat Users List' > Subject: RE: Tomcat not sending session cookies to IE > > > Ian, > > I actually have the same exact problem, except the scenario > only seems > to occur when the hostname specified in the request is that > for a cluster. > If I use the IP it works fine. It also works fine if I access > the actual > cluster machines, whether I specify the IP or the hostname. > The only time, > the session is terminated is when I specify the cluster > hostname and in > that scenario, I have the exact symptoms you describe. > > This probably doesn't help you, but at least you know you're > not alone! > If you happen upon a solution, I would be grateful for an > e-mail. Likewise, > I will let you know if we find a solution. > > Mike > > -----Original Message----- > From: Ian Bruseker [mailto:[EMAIL PROTECTED] > Sent: Friday, March 14, 2003 6:04 PM > To: Tomcat Users List > Subject: Re: Tomcat not sending session cookies to IE > > > On Friday 14 March 2003 9:18 am, Dennis Cartier wrote: > > Ian, > > > > If you are stumped, you could try running a sniffer to see > just what is > > included in the response. Set-Cookie header or not. Tcpdump > on any Unix > box > > or Ethereal comes in handy for this. > > > > Dennis > > > Well first off, let me thank you for introducing me to such a > cool new toy. > I > installed Ethereal and did some sniffing, and there's just a > whole lot of > cool stuff you can find out with that program. What a fun thing for a > Friday > afternoon. :-) > > Anyway, back to the issue at hand. I did two runs with > Ethereal, one from > the > test box browser to the client server Tomcat, which does not > work, and then > from the client server IE to the test box Tomcat, which does > work. Tomcat > is > sending the cookie in both instances, but the next connection > from the > client, it does not send the cookie back from the test box > browser for the > client server Tomcat (it does in the other case). So, why is > the browser > rejecting that cookie? I tried one last thing (which I > should have tried > long ago) - I used the IP address of the machine in the URL > instead of its > name. And guess what? I works. So it seems those who said > it was a domain > > thing were right, but I can't figure out exactly why it's > acting this way > when Mozilla didn't have a problem using the hostname. > > The hunt for the perfect answer continues (but at least now it sort-of > works). > > Ian > > > --------------------------------------------------------------------- > 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] > >
