Hello all,
I have this topology (http://alexm.unetvale.com.br/squid.png), and the
settings below from my squid.conf.
'Everything is working fine', but sometimes the load of the cpu is
very high, and I can't find where is the problem.
I suspect from the fact that's I have 2 large cache_dir on each
process, and this is generating the delay.
Today I had to format all caches_dir, and so far I have not this problem.
My question is, Can squid slow down, when used with large cache_dir,
and they are close to 80% of use?
/CLUSTER1 and /CLUSTER2 are separeted hard-drivers - (sas 15K) - just
for cache, no system files, and fs type is ext2.
I have close to 400 requests/second on Process2, and 200
requests/second on Process1.
Best regards
-------
squid.conf.1
------
wccp2_weight 700
tcp_outgoing_address __IP_ADDRESS_TELCO_B__
cache_dir aufs /CLUSTER1/cache/nucleo1/ 25396 32 256
cache_dir aufs /CLUSTER2/cache/nucleo1/ 15360 24 256
--------
squid.conf.2
-------
wccp2_weight 1100
tcp_outgoing_address __IP_ADDRESS_TELCO_C__
cache_dir aufs /CLUSTER1/cache/nucleo2/ 25396 32 256
cache_dir aufs /CLUSTER2/cache/nucleo2/ 15360 24 256
Alex Montoanelli
Administração de Redes | Desenvolvimento
Unetvale Conectividade
alex_at_unetvale.com.br
Received on Wed Mar 31 2010 - 01:49:17 MDT
This archive was generated by hypermail 2.2.0 : Wed Mar 31 2010 - 12:00:06 MDT