I run a gnetcache/gwebcache on my server here, recently due to an upgrade it was broken -- it lost access to the data files and would spit out nothing but errors about that.

All of the other clients saw the invalid data, ignored it and went on. XoloX on the other hand just kept hammering my server. When it recieved a bad answer it'd turn around immediately and hit the server again. I've got one host with over 100k hits because of this, and others not far behind him.

This needs to be fixed. If needed/wanted I can easily reproduce the server side problem that caused this but XoloX needs to just give up if it gets bad data from a cache.

There's the possibility that this has somethign to do with people running out of resources, for the time period the gcache was not behaving I have ~40 (or more) unique addresses attempting to hammer the webserver for an answer.