RE: 2.4DEVEL -> 2.4STABLE?

> -----Original Message-----
> From: Henrik Nordstrom [mailto:hno@hem.passagen.se]
> Sent: Friday, 20 October 2000 5:38 PM
> To: Adrian Chadd
> Cc: squid-dev@squid-cache.org
> Subject: Re: 2.4DEVEL -> 2.4STABLE?
>
>
> Adrian Chadd wrote:
>
> > Thing is, squid -S wasn't documented in the usage() when I
> did my initial
> > modio work which is why I added it. Hrm, I wonder who
> actually DID add it
> > and whether I just screwed up when I broke th code out into
> per-storeFS..
>
> IIRC It is a diganostic tool for checking the internal consensisty of
> the cache, only meant for verifying that the normal cache validation
> performed on startup does what it is supposed to be doing.
>
> I don't think it was ever meant that this option should be used by
> mortals...
>
> /Henrik
>
>
Strange then that the normal validation on startup code doesn't do
anything...

I suggest that

the doublecheck run automatically on a dirty startup and
-S forces a doublecheck and
the doublecheck actually fix the errors it finds (as per my tiny patch).

perhaps allow -S reportonly or something to restore the assert()
behaviour and not (touch the swap.state contents or remove files).

That should cover all bases.

Rob
Received on Fri Oct 20 2000 - 01:00:26 MDT