On Aug 5, 2009, at 10:03 AM, Henrik Nordstrom wrote:
ons 2009-08-05 klockan 07:58 -0700 skrev Ric:
I've recently been made aware that Apache supports something called
"end-to-end reload" (apparently documented in http://www.w3.org/Protocols/rfc2616/rfc2616-sec14.html#sec14.9.4)
.
I'm curious if Squid exhibits similar behavior.
Yes, by default unless you override it.
The way the Apache developers appear to interpret this spec, a
*request* containing a "Cache-Control: no-cache" header will
essentially "purge" the Apache's cache during a forced reload request
to the backend. Previously I had assumed that a client-initiated
forced reload would just bypass any proxy caches without touching the
cached entries themselves, except for the client's cache. Is this
not
the way Squid behaves?
Well, while it is true that the no-cache request in itself do not
touch
the cache Squid still acts on the response.
Regards
Henrik
Yes, that is what I meant. Squid passes the request through, leaving
it's cache untouched. Apache on the other hand appears to expire it's
cache in the process (although I have not yet tested this).
I've re-read the relevant spec a few times and I'm still not sure
which behavior is the correct one. Any thoughts?
Ric
