tuning/configuration advice

Hi guys,

I need help/advice/suggestion/comments for my squid box.

We have 128kbps upstream link (which is sometimes get worse 'coz the
physical media using 3 hop microwave transmissions) and using 2 squid
box using HP Netserver LX series.

The upstream link already congested, so i can't figure it out whether
those squid box already well-tuned or actually can achieve better
performance.

Included below are some portions of dmesg and squid.conf from squid box
we're using.
[If there are any other statistical information needed, please mention
so I can provided here]

part of dmesg output:
------------------
...
FreeBSD 3.4-RC #3: Mon Dec 13 13:40:14 JAVT 1999
    denisz@cache1.telkomsel.co.id:/usr/src/sys/compile/CACHE12
...
CPU: Pentium Pro (199.43-MHz 686-class CPU)
  Origin = "GenuineIntel" Id = 0x619 Stepping = 9
 
Features=0xf9ff<FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,SEP,MTRR,PGE,MCA,CMOV>
real memory = 805306368 (786432K bytes)
avail memory = 780726272 (762428K bytes)
...
da0 at ahc0 bus 0 target 0 lun 0
da0: <HP 4.26GB A 80-5151 5151> Fixed Direct Access SCSI-2 device
da0: 40.000MB/s transfers (20.000MHz, offset 8, 16bit), Tagged Queueing
Enabled
da0: 4067MB (8330543 512 byte sectors: 64H 32S/T 4067C)
da2 at ahc0 bus 0 target 3 lun 0
da2: <HP 9.10GB A 80-5151 5151> Fixed Direct Access SCSI-2 device
da2: 40.000MB/s transfers (20.000MHz, offset 8, 16bit), Tagged Queueing
Enabled
da2: 8678MB (17773524 512 byte sectors: 64H 32S/T 8678C)
da1 at ahc0 bus 0 target 1 lun 0
da1: <HP 9.10GB A 80-5151 5151> Fixed Direct Access SCSI-2 device
da1: 40.000MB/s transfers (20.000MHz, offset 8, 16bit), Tagged Queueing
Enabled
da1: 8678MB (17773524 512 byte sectors: 64H 32S/T 8678C)
...
--------------------

both da1 and da2 disk configured for serving cache object storage.

squid.conf:
--------------------
...
cache_peer bo.cache.nlanr.net parent 3128 3130 round-robin no-query
cache_peer sd.cache.nlanr.net parent 3128 3130 round-robin no-query
cache_peer 10.1.83.6 sibling 8080 3130 proxy-only
cache_mem 512 MB
cache_dir /data1/.squid.cache 7680 16 256
cache_dir /data2/.squid.cache 7680 16 256
dns_children 10
...
--------------------

both squid currently using Version 2.2.STABLE5-hno.19991212 with
--enable-snmp switch
(yes, I currently trying the one with patches from Henrik Nordstrom
<hno@hem.passagen.se> ;)

The squid box, however, frequently generates this error in its
cache.log, e.g:
...
1999/12/17 14:25:58| TCP connection to sd.cache.nlanr.net/3128 failed
1999/12/17 14:27:15| TCP connection to bo.cache.nlanr.net/3128 failed
...

I don't know whether it was generated because of dropped packet from our
bad lines or the squid were 'too sensitive' ;)

I'm including some daily stats gathered using mrtg here (the squid up
since last ~2 days)
[all stats presented in max/avg/current style]

- Client traffic volume (Out): 852.0/201.0/116.0 kb/min
- Client traffic volume (In): 55.0/13.0/4.0 kb/min
- Memory Usage daily: 1082.0/623.0/1082.0 kb
  (last 'saturated value is 5082.0 kb using 2.2.STABLE5)
- Swap Size daily: 14.2/14.2/14.2 M
- HTTP requests daily: 102.0/25.0/13.0 req/min
- HTTP Hits daily: 32.0/5.0/6.0 req/min

I really appreciate your comment/advice.

Regards,
-denisz-
Received on Fri Dec 17 1999 - 01:49:24 MST