Today the default setting in our DCRUM configuration is not to 'Report URL after redirect'.
We also do not use URL autolearning, due to the way the environment is set up : we have a number of different web applications exposed to the outside world all running on the same jvms; some are heavily used, other less but we still want to know the exact usage for each application (autolearning might cause us to miss the lesser used).
In the Software Service definition for the webserver tier all possible urls are defined, some refer to the same Pagename (ie. are redirected). Usage reporting is based on this Pagename (defined in CAS Business Units).
If I now change the option to 'Report URL after redirect', will this mean that I can safely remove the urls that are redirected from the Software Service definition without loosing any data?
The “Report URL after redirect” flag is used to decide which URL should be used to report the page: the base hit (HTML) URL or the first redirect that started the load of it.
If it is enabled, it means that the base hit URL is preferred and used to report the page.
Does it mean you could safely remove all user-defined URLs corresponding to “redirects”? No. In such case, you will lose information about orphan redirects (unless it is acceptable).
Also, in case you have a user-defined URL for both the base hit and the redirect (rather rare case, but possible) - if you change the flag, the base hit URL will be reported instead of the redirect URL.
By the way, if you are interested in monitoring all loads, all URLs, have you considered setting the “URL auto-learning” to “All” instead of “Off”?
As i see it, turning on the Auto - Learning would work as you say, but only until the pool is filled. After that only the highest volume URL's will be reported, so as to keep the pool filled with the busiest URL's rather than have them all.