Re: [squid-users] Transparent Squid Stalls For Up To Two Minutes

Doug Eubanks wrote:
> I'm having an intermittent squid issue. It's plagued me with CentOS 5.x, Fedora 6, and now Fedora 11 (all using the RPM build that came with the OS).
>
> My DD-WRT router forwards all of my outgoing port 80 requests to my transparent proxy using IP tables. For some reason, squid will hang when opening a URL for up to two minutes. It doesn't always happen and sometimes restarting squid will correct the problem (for a while). The system is pretty hefty 3ghz P4 with 2G of RAM with a SATA II drive. That should be plenty for a small home network of about 10 clients.
>
> When I test DNS lookups from the host, requests are returned within less than a second. I'm pretty sure that's not the problem.
>
> Here is my squid.conf, any input would be greatly appreciated!
>
> acl manager proto cache_object
> acl localhost src 127.0.0.1/32
> acl to_localhost dst 127.0.0.0/8
> acl localnet src 10.0.0.0/8 # RFC1918 possible internal network
> acl localnet src 172.16.0.0/12 # RFC1918 possible internal network
> acl localnet src 192.168.0.0/16 # RFC1918 possible internal network
> acl SSL_ports port 443
> acl Safe_ports port 80 # http
> acl Safe_ports port 21 # ftp
> acl Safe_ports port 443 # https
> acl Safe_ports port 70 # gopher
> acl Safe_ports port 210 # wais
> acl Safe_ports port 1025-65535 # unregistered ports
> acl Safe_ports port 280 # http-mgmt
> acl Safe_ports port 488 # gss-http
> acl Safe_ports port 591 # filemaker
> acl Safe_ports port 777 # multiling http
> acl CONNECT method CONNECT
> http_access allow manager localhost
> http_access deny manager
> http_access allow localnet
> http_access deny !Safe_ports
> http_access deny CONNECT !SSL_ports
> http_access allow localnet
> http_access allow localhost
> http_access deny all
> htcp_access allow localnet
> htcp_access deny all
> http_port 3128 transparent

Is the NAT / REDIRECT/DNAT happening on the Squid box?
It needs to.

> hierarchy_stoplist cgi-bin ?
> cache_mem 32 MB
> maximum_object_size_in_memory 128 KB
> cache_replacement_policy heap LRU
> cache_dir aufs /var/spool/squid 4096 8 16

4GB of objects under 512KB small (avg set at 64KB later), using only an
8x16 inode array. You may have a FS overload problem.

Also, Squid 'pulses' cache garbage collection one directory at a time.
Very large amounts of files in any one directory can slow things down a
lot at random times.

It's generally better to increase the L1/L2 numbers from default as the
cache gets bigger.

> max_open_disk_fds 0
> minimum_object_size 0 KB
> maximum_object_size 512 KB
> access_log /var/log/squid/access.log squid
> refresh_pattern ^ftp: 1440 20% 10080
> refresh_pattern ^gopher: 1440 0% 1440
> refresh_pattern (cgi-bin|\?) 0 0% 0
> refresh_pattern . 0 20% 4320
> visible_hostname doug-linux.dougware.net
> unique_hostname doug-linux.dougware.net
> coredump_dir /var/spool/squid
> cache_mgr admin_at_dougware.net
> dns_nameservers 10.0.0.254 10.0.0.253 69.197.163.239
> store_avg_object_size 64 KB
> memory_replacement_policy heap LRU
> tcp_outgoing_address 10.0.0.254
> udp_outgoing_address 10.0.0.254

Does 10.0.0.254 port 53 have access to ALL the DNS servers: 10.0.0.254
10.0.0.253 69.197.163.239

Are you excluding 10.0.0.254 from the interception at the DD-WRT?

Amos

-- 
Please be using
   Current Stable Squid 2.7.STABLE6 or 3.0.STABLE15
   Current Beta Squid 3.1.0.7