Hi all, I have been helping develop a community based website for sharing audio samples. The site now gets more traffic than we have bandwidth for, and we have been hoping to enlist some fellow academic institutions to help us mirror our files. But rather than simply transferring all the files to these mirrors, I had been thinking about placing a Squid server on each mirror, configuring it as an accelerator, and then redirecting traffic to these squid servers from the main site. Then, all traffic would pass through the Squid servers, and the most heavily requested files would be cached by the Squid servers while less frequently requested files would be pulled from the main server. Um... I'm writing to ask whether anyone thinks I will encounter any unexpected problems if I try to do this. As far as I can tell, other people have been using Squid as an accelerator to take the load of dynamically generated websites, but no one seems to be using it for file mirroring. Is there a deeper reason for this?
Best, Greg Kellum
