Selenium ignores base url

by jenny on 13 January 2010 - 06:38pm in

I usually record a test on a dev machine, then replay it on the production machine. Sometimes specifying the production machine as the base url just "doesn't work". The base url gets ignored and Selenium opens up my dev machine and proceeds to replay the test there. A bit of googling shed some light on the situation.

Grepping through my test cases, I see output like this:

$ grep '<link' *
test1:<link rel="selenium.base" href="http://dev.example.com/" />
test2:<link rel="selenium.base" href="" />

...sometimes, the dev machine gets stored in the test and sometimes it doesn't. I haven't been able to figure out why this happens intermittently. At any rate, removing all references to the dev machine and leaving an empty href attribute (as in the second line) does the trick.

Thanks

Thanks for the tip. Worked for me.