Re: Providing continuity of service during cache maintenance.

What I wouldn't mind seeing is the network layer of Squid decoupled from the
caching layer:

Essentially, by running it in two components, the cache-unit could be switched out,
or do other things (maintenance, cleanup, rebuild) while the network layer was
still passing requests through to parent(s). As a two-process setup
(squid-comms+squid-cache) we'd also wind up with a few extra file-descriptors to
play with, and each unit could be restarted independantly of the other (ease of
maintenance).

IMO, anyways. I don't see any holes in that offhand.

D

Michael James wrote:

> Could Squid be fitted with a command line flag to run in "dummy" mode;
> continuing to provide a proxy service but without caching?
>
> We run Squid-1.1.18 on a Solaris 2.6 ultra. Direct access to port 80 off
> campus is blocked.
> Recently we had trouble with cache corruption and boobs were popping out of
> normally staid sites.
>
> With 7 Gig of cache it took time to flush it and would have taken much
> longer to clean it.
> (I estimate it costs us $100-$200 to loose the data so cleaning is interesting)
>
> While I was rebuilding the main cache partition I started another squid
> with a config file
> that gave it 20 Meg on another disk.
> It would be easier to have a flag or interrupt to tell squid to run without
> the cache partition.
>
> Also ./squid -z is a gotcha as it builds the cache directory structure only
> root writeable,
> and we run squid as daemon. This could be fixed by simply including a
> warning in squid.conf
> near the "cache_effective_user" parameter.
>
> michaelj

--
Note to evil sorcerers and mad scientists: don't ever, ever summon powerful
demons or rip holes in the fabric of space and time. It's never a good idea.
ICQ UIN: 3225440