|
Obviously, something has changed in 1.572 when it comes to Jenkins generating URLs referring to itself, as explained here:
https://wiki.jenkins-ci.org/display/JENKINS/Jenkins+says+my+reverse+proxy+setup+is+broken
In practice, of course, the "Jenkins URL" parameter should be properly set. Yet in cases where there is more than one possible URL to reach Jenkins (for whatever reasons one can imagine) Jenkins still worked 99% for us even if we accessed it using a different URL than the one in that parameter.
As that behaviour has changed, for situations where there is a good reason to keep more than one possible URL, wouldn't a 2nd field there "Jenkins URL aliases" with out without redirecting make sense. (Redirect = when accessed through an alias, redirect to the one and only URL / other model: When accessed at www.alias2.url/jenkins make self-referring links also point to www.alias2.url/jenkins.
Bonus question:
In http://stackoverflow.com/questions/11221238/what-does-jenkins-url-means-in-configuration-settings in one of the answers it says:
"Jenkins can't determine its URL on its own"
Actually, why not? Besides doing a hostname and port lookup (which indeed can lead to wrong results) shouldn't the URL used to access the page be available in the HTTP headers for any one request?
|
