On 21/02/2017 10:00 p.m., FLASH FLASH wrote: > this is part of my cache.log: > > 2017/02/21 10:54:47.908 kid1| 11,2| http.cc(2221) sendRequest: HTTP Server > REQUEST: > --------- > GET /update/repair_sigver/1.0.25.18/win32/int/repair.rdf.lz HTTP/1.1 > Host: personal.avira-update.com > User-Agent: @AUVI@1.3;AntiVir-NGUpd/15.0.24.143 (PERS; WKS; EN; AVE > 8.3.42.182; VDF 8.12.155.244; Windows 10 Enterprise; ; Bulgaria; > 4e083d762df61e49122f00e3758c65355687472f; 0000149996-AVHOE-0000001; BG; > BUILD 15.0.24.146; 0; 1; 1; iexplore; 1; > 309afc5503fe48febffe89e27e9390322a9f97f2; 0) > Accept: */* > Via: 1.1 router2 (squid/3.5.10) > X-Forwarded-For: ... > Cache-Control: no-cache, no-store, must-revalidate There we go, the AV updater client (or maybe that 'router2' proxy?) forbids cached content to be delivered as a response AND the response to this request from being cached. Which is kind of annoying since the server indicates the reply is actually cacheable for at least a few minutes and HTTP/1.1 revalidation performance features are supported. :-( If router2 is under your control you could optimize a bit by erasing that Cache-Control header from the AV updaters request before they are sent on to the current Squid. Amos _______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
