mån 2009-09-14 klockan 15:08 +0100 skrev Genaro Flores: > Ah, thanks. I reckon there have been no changes since I last asked the > question. Although 'purge' does look promising (except for efficiency and > that I doubt it can do anything to decrease the time required to perform > multiple queries). This is because there is no searchable index with the data you look for. So each object has to be opened to find the exact properties. The Squid cache index is keyed on an MD5 hash of the URL (16 bytes "random" data), not even preserving the requested hostname in a meaningful manner. Regards Henrik
