On 8/06/2016 5:08 a.m., Adam Vollrath wrote: > Good morning! I'm having trouble with upstream HTTP servers not > responding for more than 30 seconds. I'd like to just return an empty > response to the downstream client. Nothing is better than waiting. > > I'm looking for a `squid3` configuration directive to timeout between an > HTTP request and response. Here's the directive for `nginx`: > http://nginx.org/en/docs/http/ngx_http_proxy_module.html#proxy_read_timeout > > "Defines a timeout for reading a response from the proxied server. The > timeout is set only between two successive read operations, not for the > transmission of the whole response. If the proxied server does not > transmit anything within this time, the connection is closed." > > Does `squid3` have a way to implement this? Is this what the > `read_timeout` directive does if my upstream systems are configured as > `cache_peer`?
I think you misunderstand what "proxy" means. Nginx primary design focus as a web server like Apache. So proxied traffic is a special activity for it. Squid is a proxy. So by definition everything going through Squid is proxied. <http://www.squid-cache.org/Doc/config/read_timeout/> controls the timeout for individual read(2) operations on a connection which is active. <http://www.squid-cache.org/Doc/config/write_timeout/> controls the timeout for individual write(2) operations on a connection which is active. <http://www.squid-cache.org/Doc/config/server_idle_pconn_timeout/> controls the timeout for connections in idle state between requests. Other timeouts control specific actions like DNS, route selection, connect(2), or client connection activities. HTH Amos _______________________________________________ squid-users mailing list [email protected] http://lists.squid-cache.org/listinfo/squid-users
