Re: RAID or multiple cache_dir s?

From: Oskar Pearson <oskar@dont-contact.us>
Date: Thu, 31 Dec 1998 12:27:44 +0200

Hi

> I was about to purchase a new box for squid (my old 2Gb, P90 has served me
> well for a year running squid 1.somthing). I spec-ed out a Dell 2300
> with 512Mb, 6x9Gb + RAID-0 (Dell PERC-RAID is really AMI MegaRaid which is
> supported in 2.0.36 linux kernels (and Redhat 5.2)).

If I was going to use lots of disks I would go raid-4 or raid-5... not
raid-0.

I would only use hardware-raid if I was to use the reliability features...
otherwise I would simply use two filesystems (or even software md-raid).

> o For a reasonably well used cache, is using multiple cache_dir entries and
> efficient as a hardare RAID-0 box? (Yes I know there will be some
> small performance loss, but will squid make use of all the disks
> efficiently?)

It will probably be as efficient. Squid balances between multiple disks.
using round-robin to write different requests to different disks.

> o Another thread asked the question if a disk (and thus a cache_dir) went
> bad would squid continue to perform well using the remaining disks ?
> Is this true?

No.... or it depends.

If your kernel code crashes because the ext2 filesystem goes bad, then no.
If your kernel survives, you may get:

Corrupted objects
0-length returned objects (because Squid cannot open files, cannot read
        from them or something similar)
Wrong objects returned.

> o Whats the maximum limits on Squid? Should I continue to throw disk at a
> box, or buy a separate box. If I can save on RAID, I can afford extra
> memory and so have a 1Gb RAM box - Squid won't break?

There are pages out there that give 'squid sizing' on various platforms. I
suggest that you have a look around.

Remember - if you compile Squid-2* with async-io (./configure
--enable-async-io) then you will be able to exceed performance compared to
the same hardware running 1.0 and 1.1.

> o Is it the general consensus to avoid Transparent proxying if possible, and
> that it is better to force the user to knowingly use the proxy settings in
> their browser?

Be careful. Make sure that all managers/customers/etc know before hand.
You could have problems with people that need direct access (think about
people using IP address authentication).

I would use transparency here: we have too large a user base for all people
to request help changing their settings. If we were to do it, though, we
would do it very, very, very carefully.

Oskar
Received on Thu Dec 31 1998 - 03:16:13 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:43:46 MST