A couple of pconn questions;
1) Squid supports HTTP/1.0-style persistent connections; i.e., if it
gets a request with a Connection: keep-alive header in it, it will
reuse the connection.
However, if it receives a HTTP/1.1 request, it will fall back to one-
request-per-connection. Since pconns are the default in HTTP 1.1, why
not use them?
In client_side.c, line 597 (clientSetKeepaliveFlag);
---8<---
httpBuildVersion(&http_ver, 1, 0); /* we are HTTP/1.0, no matter
what the client requests... */
if (httpMsgIsPersistent(http_ver, req_hdr))
request->flags.proxy_keepalive = 1;
}
--->8---
would need to be
---8<---
if (httpMsgIsPersistent(request->http_ver, req_hdr))
request->flags.proxy_keepalive = 1;
}
--->8---
httpMsgIsPersistent already appears to do the right thing. Are there
any particular reasons why this is disabled? While Mozilla and IE
still send Connection: keep-alive with 1.1 requests, Safari does not...
2) Squid still sends Proxy-Connection headers. As far as I can see,
they're not required by any modern implementations; everybody
understands Connection. Maybe it's time to stop generating them?
3) Squid can't persist client-side connections if it doesn't have a
Content-Length handy, so if the origin server doesn't provide one,
it'll close. However, responses cached by Squid -- by their very
nature -- have a C-L available to Squid, even if the origin server
doesn't send one. Since content generated by scripts often doesn't
have C-L set, but can sometimes be cacheable, it would be a nice
optimisation to synthesise the response body length if you don't have
a C-L on a cached response. Has anyone attempted this?
Cheers,
--
Mark Nottingham
mnot@xxxxxxxxxxxxx
