[squid-users] Squid and larger environments

Hello,

I was wondering if there were any configuration recommendations for
Squid in large environments. We have approximately 4000+ users.
Below is my squid configuration as it stands now without the acl and
cache_peer_access (and a few other) lines. System configuration is
below that. I know this is a dual CPU and squid does not really take
advantage of this configuration. Also, after reading a few of the
posts here I have been seriously considering removing the RAID1
configuration on the drives.

http_port 3128
acl QUERY urlpath_regex cgi-bin \?
acl SSL method CONNECT
acl PURGE method PURGE
no_cache deny QUERY
cache_mem 768 MB
max_filedesc 8192
maximum_object_size 1024 KB
maximum_object_size_in_memory 128 KB
ipcache_size 4096
fqdncache_size 4096
cache_dir aufs /var/spool/squid_cache 4000 16 256
#cache_dir null /tmp
cache_store_log none
cache_access_log /var/log/squid/access.log
logfile_rotate 60
#debug_options ALL,9
quick_abort_min 16 KB
quick_abort_max 8192 KB
quick_abort_pct 50
negative_ttl 1 minutes
positive_dns_ttl 1 hours
negative_dns_ttl 1 minutes
range_offset_limit 1024 KB
shutdown_lifetime 10 seconds

memory_pools off
query_icmp off
test_reachability off
buffered_logs on
never_direct allow all
snmp_port 3401
snmp_access allow snmppublic localhost
snmp_access allow snmppublic mgmt
snmp_access allow snmppublic proxy_net
snmp_access deny all
max_open_disk_fds 0
uri_whitespace allow
nonhierarchical_direct off
prefer_direct off
strip_query_terms on
coredump_dir /var/spool/squid
ignore_unknown_nameservers off
client_persistent_connections on
server_persistent_connections on
half_closed_clients on
pipeline_prefetch on
request_entities on
ie_refresh on
forwarded_for off

HP DL-380 G4
Dual 3.2GHz CPU
4GB Memory
2x72GB Drives - RAID 1 - Operating System
2x146GB Drives - RAID 1 - Cache

If more information is needed, please let me know.

Thank you,
Troy
Received on Mon Feb 12 2007 - 12:50:28 MST