[squid-users] Persistent Server connections, pipelining and matching responses
Hi: I am trying to understand Squid behavior when server side connection is pinned (persistent) and pipelining is enabled on both client and server side in a transparent proxy configuration. If there are multiple HTTP requests coming from multiple clients for the same server and if the requests are sent on the same TCP connection to the server, how will the proxy match the responses to those requests to appropriate clients ? HTTP is stateless - so the response will not identify the request. If the proxy gets the reply to all those requests, how will it send the right response to the right client. Thanks in advance. -- View this message in context: http://squid-web-proxy-cache.1019090.n4.nabble.com/Persistent-Server-connections-pipelining-and-matching-responses-tp2540989p2540989.html Sent from the Squid - Users mailing list archive at Nabble.com.
Re: [squid-users] Persistent Server connections, pipelining and matching responses
I do not believe squid uses a single TCP Persistent Connection for more than 1 client at a time, hence the functionality of Connection Pinning. This is required to perform things such as NTLM web-based logins. - Chad E. Naugle Tech Support II, x. 7981 Travel Impressions, Ltd. cachenewbie email2sr...@yahoo.com 9/15/2010 2:27 PM Hi: I am trying to understand Squid behavior when server side connection is pinned (persistent) and pipelining is enabled on both client and server side in a transparent proxy configuration. If there are multiple HTTP requests coming from multiple clients for the same server and if the requests are sent on the same TCP connection to the server, how will the proxy match the responses to those requests to appropriate clients ? HTTP is stateless - so the response will not identify the request. If the proxy gets the reply to all those requests, how will it send the right response to the right client. Thanks in advance. -- View this message in context: http://squid-web-proxy-cache.1019090.n4.nabble.com/Persistent-Server-connections-pipelining-and-matching-responses-tp2540989p2540989.html Sent from the Squid - Users mailing list archive at Nabble.com. Travel Impressions made the following annotations - This message and any attachments are solely for the intended recipient and may contain confidential or privileged information. If you are not the intended recipient, any disclosure, copying, use, or distribution of the information included in this message and any attachments is prohibited. If you have received this communication in error, please notify us by reply e-mail and immediately and permanently delete this message and any attachments. Thank you.
Re: [squid-users] Persistent Server connections, pipelining and matching responses
ons 2010-09-15 klockan 11:27 -0700 skrev cachenewbie: I am trying to understand Squid behavior when server side connection is pinned (persistent) and pipelining is enabled on both client and server side in a transparent proxy configuration. In default configuration Squid serializes pipelined requests, processing them one at a time. If there are multiple HTTP requests coming from multiple clients for the same server and if the requests are sent on the same TCP connection to the server, how will the proxy match the responses to those requests to appropriate clients? Squid just sends one request at a time per server connection, and reuses the same connection for another request when the complete response have been seen. In future we may pipeline many requests concurrently under specific conditions. HTTP defines how replies match up with their requests by same order. If requests A B C is sent pipelined to a server then the server MUST respond in the same order, response A B C. Regards Henrik
