Forum Discussion
BIG-IP : redirect loop : debugging techniques
BIG-IP 11.4.0 Build 2384.0 Final
In my hosts file I point www.mydomain.com to my web-server IP, and in my client-browser I submit the request
http://www.mydomain.com/natural-language-url-1 ...
nitass_89166
Noctilucent
Noctilucentyou want curl from f5 command line. But how to page the output ? because it's overflowing the putty window - even with |more
you can use -I to request only header (i.e. HEAD method).
e.g.
curl -Iv -H "Host: www.mydomain.com" http://ip_address/natural-language-url-1
based on the output, requests seem to be same but responses are different. is there any specific configuration on server?
would it be possible to not use snat as John asked?
I am not a networking engineer, but here is my understanding : This is not an isolated test environment, and to construct one is prohibitively complex & time-consuming. The dev-test F5 ( which I am using to develop the new iRules ) is on the same internal engineering subnet as other non-prod F5s. To be functional, the dev-test website must talk to shared db-servers and web-services, and therefore I have the web-server gateway set to a VIP of one of those other F5s ( otherwise I would have to do a total network configuration on the dev-test F5 - as opposed to just what I need to test natural-language URLs ). I added SNAT to the dev-test F5 virtual-server because the only other option to ensure responses flowed back through the F5 was to add a static route to the web-server, and this was not representative of how it would work in a prod environment. One thing I have noticed is that some of the virtual addresses ( different from those used by the vips the test harness hits ) are red, indicating they are not available, but I don't know why. I am wondering if one of these vips could be the self-ip and this is the underlying cause of the redirect loop error ?