Its all fixed now. For those who care here is what happened:
I was using caching by saving the xml output from flickr’s api. Then what SHOULD have happend is this, if there was a cache-hit, it was supposed to read from that file. WELL, as you may know – php’s file_get_contents allows you to read from a url as well as a local file… I didn’t notice that I was calling the API EVERY time. If the flickr staff wasn’t at etech and SXSW, they would have shut me down much sooner!