tor 2010-09-09 klockan 09:25 -0500 skrev Jorge Iván Burgos Aguilar: > Dump question for anyone with experience in building redirectors: > How does concurrency is being implemented in the redirectors... > A) A parallel-connection to the redirector > B) More than one line at once > C) Both > ??? B and also D (multiple processes). The concurrency protocol sends more than one request at a time to the the helper, each tagged with a unique id. The helper may respond to the queries in any order if needed. http://www.squid-cache.org/Doc/config/url_rewrite_program/ http://www.squid-cache.org/Doc/config/url_rewrite_concurrency/ Same scheme applies to all helper channels supporting concurrency. In addition to concurrency there is also the children parameter which controls how many instances of the helper Squid starts. Regards Henrk
