[squid-users] Squid workers on non cache dir rock system
Dave Dykstra
dwd at cern.ch
Mon Jan 13 22:35:49 UTC 2025
On Mon, Jan 13, 2025 at 02:17:01PM -0500, Alex Rousskov wrote:
> On 2025-01-13 13:29, Jonathan Lee wrote:
...
> > Is there anyway to use more workers on a non rock system, without
> > disabling the cache? I can use them when cache is disabled. Without
> > it I get assertion failed: controller:cc:930: EX"
...
> 3. Using multiple workers with non-rock cache_dir is not supported. For
> some definitions of "work", in some deployment environments, one may be
> able to use such unsupported configurations (e.g., by using
> configuration macros to create worker-specific cache_dirs), but I do not
> recommend those unsupported tricks and am not going to detail them here.
> Sooner or later, those tricks may stop working without notice and with
> bad side effects.
But in case you want to try it anyway, see the ${process_number} macro.
https://github.com/squid-cache/squid/blob/v6/src/cf.data.pre#L123
It has so far worked for me with ufs cache and enabled a far greater use
of full machines without the need for rock cache, which I can't use due
to https://bugs.squid-cache.org/show_bug.cgi?id=4890.
Dave
More information about the squid-users
mailing list