Re: Is squid unable to handle the load? from Brian on 1998-05-27 (squid-users)

From: Brian <signal@dont-contact.us>
Date: Wed, 27 May 1998 08:49:08 -0500 (CDT)

On Tue, 26 May 1998, Mark Dabrowski wrote:

> Ok, here's what I've done:
>
> We have ordered 3 new Adaptec 2940 U2W controllers (UltraWide 2) which go
> 80MB/s, but unfortunately couldn't find anywhere 4GB Ultra Wide 2 drives
> and also BSD 3.1 does not recognize (support) 2940U2W at all ...
>
> So we ended up installing 2 more 2940 UW's and adding 2 x 4GB drives.
>
> The disk configuration of Squid box is 3 AHA 2940 UW controllers and 1 X
> 9GB Drive and 2 X 4.5GB drives. I have defined 3 equal partitions in
> cache_dir, each on each disk.
>
> Since we just finished installation, i didn't have a chance to make any
> tests, but when the system is almost idle and only 5-10 clients use squid
> there is something weird happening... Every 15-20 seconds I would notice
> number of disk operations/sec jump to about 100 on each disk (which causes
> cpu idle to go down to 100%) and then resumes normally.
>
> Is squid doing some disk operations every 15-20 seconds? If yes, under
> normal conditions I would not think it can halt the system.... but it
> does... Any ideas?

Did you stripe the drives with RAID0? I personally recommend a 4*2GB
configuration, RAID0 striped.

Brian

>
> procs memory page disks faults cpu
> r b w avm fre flt re pi po fr sr s0 s1 s2 in sy cs us
> sy id
> 0 2 0 129396 199540 5 1 1 1 0 1 3 3 3 205 536 41 2
> 2 96
> 1 2 0 129396 199540 5 1 1 1 0 1 0 3 4 228 795 63 0
> 6 94
> 1 2 0 129396 199540 5 1 1 1 0 1 6 7 3 289 1242 78 4
> 6 90
> 0 2 0 129396 199548 6 1 1 1 0 1 3 4 4 299 1383 89 1
> 11 88
> 1 2 0 129396 199556 5 1 1 1 0 1 8 4 6 298 1255 83 4
> 10 86
> 0 2 0 129936 199556 5 1 1 1 0 1 0 6 0 226 681 57 3
> 3 94
> 0 2 0 129936 199560 5 1 1 1 0 1 3 3 5 268 932 63 3
> 0 97
> 0 2 0 129936 199568 5 1 1 1 0 1 6 5 6 254 1010 69 1
> 8 91
> 0 3 0 130388 199568 5 1 1 1 0 1 63 108 104 481 392 33 2
> 2 100
> 0 2 0 130388 199580 5 1 1 1 0 1 81 31 15 353 788 28 3
> 5 96
> 0 2 0 130388 199580 5 1 1 1 0 1 3 3 6 210 585 48 4
> 2 94
> 0 2 0 130388 199580 5 1 1 1 0 1 0 0 0 198 384 33 0
> 1 99
> 0 2 0 130388 199580 5 1 1 1 0 1 0 1 0 188 564 46 1
> 2 97
> 0 2 0 130388 199580 5 1 1 1 0 1 3 5 0 239 734 60 3
> 2 95
> 0 2 0 130388 199572 9 1 1 1 0 1 1 0 6 202 723 51 2
> 3 95
> 0 2 0 130388 199564 7 1 1 1 0 1 3 3 3 199 420 46 0
> 3 97
> 0 2 0 130388 199564 5 1 1 1 0 1 4 3 0 203 573 56 3
> 5 92
> 0 2 0 130388 199564 5 1 1 1 0 1 0 0 3 200 606 56 1
> 2 97
> 0 2 0 130388 199564 5 1 1 1 0 1 3 0 0 183 372 36 2
> 2 96
>
>
> Mark Dabrowski
>
>
> At 12:04 5/27/98 +1200, you wrote:
> >+ Are you running ultra drives? is ultra enabled? How many Commands are you
> >+ sending to the controller simultaneously? You can genereally tweak the
> >+ 2940 quite a bit, enable tag queueing, paging, ultra, etc.
> >
> >Could be worth it but for a single disk 100 IOPS is about what you'd expect.
> >
> >Go with the earlier advice and get a couple of disks and stripe across them.
> >The bottleneck shouldn't be on the controller (unless it is quite broken), so
> >you should be able to put the disks on the same controller and get really
> good
> >performance increase.
>
>
> +------------------------------------------<<<
> + Mark Dabrowski @ Canada On-Line @ Toronto, Ontario
> + Tel: (905) 624-8566 ext.5067, Fax: (905) 624-5397, mark@ica.net
> + For technical inquires please email support@ica.net or call our
> + Technical Support Department at (905) 624-1953.
> +------------------------------------------------------------------>>>
>

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:40:21 MST