Re: Is squid unable to handle the load?

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?

 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.
+------------------------------------------------------------------>>>
Received on Tue May 26 1998 - 18:03:51 MDT