Recently, one of our accelerated machines started throwing out errors, and it got me thinking. Would it be possible to have Squid not cache objects based on a status code (for instance 500/Internal server error, 403 forbidden etc)? This would be something along the lines of content-inspection, so a quick take of the flow that would happen would be: 1: Client connects to Squid and requests www.normally.cacheable/object 2: Squid notices that the object is stale and attempts to retrieve a fresh copy from the origin server 3: Origin server returns Status: 500 in the headers, and Squid defaults to serving up the stale object instead of the fresh (but broken) object Any ideas, finger-pointing or such would be greatly appreciated / oscar
