The major use of nocache_hack is in refresh.c. There it makes a
significant difference.
On Sunday 28 April 2002 11:39, maer727@sohu.com wrote:
> Thanks, Henrik pal!
>
> I have read the codes,
> if (no_cache) {
> #if HTTP_VIOLATIONS
> if (Config.onoff.reload_into_ims)
> request->flags.nocache_hack = 1;
> else if (refresh_nocache_hack)
> request->flags.nocache_hack = 1;
> else
> #endif
> request->flags.nocache = 1;
> }
>
> I think it means that in this case nocache and nocache_hack are not
> the same. For example, in this case, nocache==true but
> Config.onoff.reload_into_ims==false, nocache and nocache_hack are
> not the same. Am I correct?
>
> In the pase time, I only notice the following two cases, so in my
> mind they are the same all the time. :-) Now I think I have known
> in some cases they are not always the same.
>
>
> if (r->flags.nocache)
> ipcacheInvalidate(r->host);
> #if HTTP_VIOLATIONS
> else if (r->flags.nocache_hack)
> ipcacheInvalidate(r->host);
> ......
> if (r->flags.nocache || r->flags.nocache_hack) {
>
> Best regards,
> George Ma
>
> ----- Original Message -----
> From: Henrik Nordstrom
> To: maer727@sohu.com
> Cc: squid-dev@squid-cache.org
> Subject: Re: My opinions on nocache and nocache_hack. :-)
> Sent: Sun Apr 28 16:43:13 CST 2002
>
> > http://www.squid-cache.org/mail-archive/squid-dev/200204/0707.htm
> >l
> >
> > On Sunday 28 April 2002 05:36, maer727@sohu.com wrote:
> > > Thanks, Henrik pal!
> > >
> > > I search "no_cache" in squid.conf and do not find any rules.
> > > What are the rules you mentioned?
> > >
> > > Another question, what means "may override the request"? Do you
> > > mean in some cases even though the request asks the cache to
> > > remove the object from cache, the object will still be kept in
> > > cache without removing(so, it is on the contrary of the
> > > request)?
> > >
> > > Best regards,
> > > George Ma
> > >
> > > ----- Original Message -----
> > > From: Henrik Nordstrom
> > > To: maer727@sohu.com
> > > Cc: squid-dev@squid-cache.org
> > > Subject: Re:Re: My opinions on nocache and nocache_hack. :-)
> > > Sent: Sat Apr 27 22:00:36 CST 2002
> > >
> > > > No, they are not the same. Both indicates the user has
> > > > requested the object must not be from the cache, but in case
> > > > of nocache_hack there is squid.conf rules which may override
> > > > the request.
> > > >
> > > > On Saturday 27 April 2002 14:20, maer727@sohu.com wrote:
> > > > > Thanks, Henrik pal!
> > > > >
> > > > > I think nocache_hack means the cache should not cache the
> > > > > object. I mean the meaning of no_cache and nocache_hack are
> > > > > the same.
> > > > >
> > > > > Am I correct?
> > > > >
> > > > > Best regards,
> > > > > George Ma
> > > > >
> > > > > ----- Original Message -----
> > > > > From: Henrik Nordstrom
> > > > > To: maer727@sohu.com
> > > > > Cc: squid-dev@squid-cache.org
> > > > > Subject: Re: My opinions on nocache and nocache_hack. :-)
> > > > > Sent: Sat Apr 27 15:43:25 CST 2002
> > > > >
> > > > > > maer727@sohu.com wrote:
> > > > > > > Hi, pals!
> > > > > > >
> > > > > > > After reading severals of source codes, I think
> > > > > > > nocache_hack is the same as nocache if HTTP_VIOLATIONS
> > > > > > > is defined. I have met with serveral cases where we
> > > > > > > treat the two variables the same if HTTP_VIOLATIONS is
> > > > > > > defined.
> > > > > > >
> > > > > > > Am I correct?
> > > > > >
> > > > > > Not entirely. nocache_hack is a special mode of dealing
> > > > > > with no_cache. See the following code segment from
> > > > > > client_side.c where the two flags are set in the request:
> > > > > >
> > > > > > if (no_cache) {
> > > > > > #if HTTP_VIOLATIONS
> > > > > > if (Config.onoff.reload_into_ims)
> > > > > > request->flags.nocache_hack = 1;
> > > > > > else if (refresh_nocache_hack)
> > > > > > request->flags.nocache_hack = 1;
> > > > > > else
> > > > > > #endif
> > > > > > request->flags.nocache = 1;
> > > > > > }
> > > > > >
> > > > > >
> > > > > > Config.onoff.reload_into_ims and refresh_nocache_hack
> > > > > > depends on your squid.conf configuration
> > > > > > (reload_into_ims, or the use of no-cache related options
> > > > > > to refresh_pattern).
> > > > > >
> > > > > > Regards
> > > > > > Henrik
Received on Sun Apr 28 2002 - 06:20:34 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:15:23 MST