On Mon, Mar 17, 2008 at 3:20 AM, Alex Rousskov
<rousskov@xxxxxxxxxxxxxxxxxxxxxxx> wrote:
> On Mon, 2008-03-17 at 06:25 +0900, Adrian Chadd wrote:
> > On Fri, Mar 14, 2008, Alex Rousskov wrote:
> >
> > > > I think it actually is a bug in the Vary handling in Squid-3.
> > > > The condition:
> > > >
> > > > if (!has_vary || !entry->mem_obj->vary_headers) {
> > > > if (vary) {
> > > > /* Oops... something odd is going on here.. */
> > > >
> > > > .. needs to be looked at.
> > >
> > > But it is not the condition getting hit according to Aurimas' log, is
> > > it?
> >
> > There's two confusing log messages with the same message, which is confusing
> > things.
>
> The messages are indeed poorly written, but they are different:
>
>
> varyEvaluateMatch: Oops. Not a Vary object on second attempt,
> varyEvaluateMatch: Oops. Not a Vary match on second attempt,
>
> That is why I said that it was the other, arguably less suspicious,
> condition being hit here.
Right, since it's the case with "Vary match on second attempt", for
now I just "moved" it to debug level 2 with a help of Alex patch.
Thanks! BTW, let me know if I can help debugging this.
// Aurimas
