Hello, only two more questions about pipelining in squid: if squid supports this feature, what happens in the following situation? - a client sends a set of consecutive requests - squid receives the first one, the required object is not in cache, so sends it to the server - the response of the server has not been still received, when squid receives the next request of the client. The required object is in cache. To ensure the packets order through the pipelined communication channel, squid would have to wait for the first server response before sending this last object, found in the cache. Is this true? Finally, if the pipelining is supported, and initially squid has an empty cache, assuming a bandwidth of X B/s for the proxy-client communication, and Y B/s for the proxy-server communication, we can say that, for a set of requested objects with a total size of M bytes, the client can obtain them in about M/Ts + 1/X + 1/Y + 1/Tp + Ti seconds (Ts = server throuput time, Tp = proxy throuput time, Ti = arrival time of the first request in server). Is this correct? Sorry about these strange questions, but are necessary for my current job. Thanks for all. Best regards, JP
