Re: [squid-users] Squid high bandwidth IO issue (ramdisk SSD)

Have you actually done any system profiling to get an understanding of
what wall(s) you're hitting?

For example, is there perhaps some strange issue you're hitting with
the disk controller?

I'm pushing around 100-150mbit at peak through one server right now
but I've not yet finished deploying COSS. But then, I'm hitting other
issues with TCP stack tuning and CPU use.

Adrian

2009/8/2 smaugadi <adi_at_binat.net.il>:
>
> Dear ALL,
> We have a squid server with high volume of traffic, 200 – 300 MB.
> The server is in transparent mode and using 18GB of ramdisk. With this
> configuration performance is very good (after optimizing the squid and the
> linux machine).
> The problem is the small size of cache directory.
> Since IO is a big issue with squid we purchased the intel ssd x25-m 160gb
> with hope that a flash drive will overcome the IO problems.
> Unfortunately we failed. After adding the SSD disk and configuring squid to
> use another cache directory we encountered high IO that led to performance
> degradation.
>
> We have tried using AUFS and even downgrading to version 2.6 and using COSS
> with no success.
> Does any one have an idea how to overcome the IO problems?
> Do you have a suggestion to configuring squid in a big enterprise with big
> traffic volume?
>
> We have searched the WEB and used as reference links such as:
> http://www.freeproxies.org/blog/2007/10/03/squid-cache-disk-io-performance-enhancements/
> http://blog.last.fm/2007/08/30/squid-optimization-guide
>
> Platform: x3455 – 2 x dual core opteron 2.5Ghz
> OS Memory: 32GB (24GB to ramdisk)
> OS: Linux RHEL 5.0
>
> Squid configuration:
>
> 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 ib src XXXXXXXXXXXXXX
> acl mgt src XXXXXXXXXXXXX
> 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 deny !Safe_ports
> http_access deny CONNECT !SSL_ports
> http_access allow ib
> icp_access allow localnet
> icp_access deny all
> htcp_access allow localnet
> htcp_access allow all
> http_port 3128 transparent
> hierarchy_stoplist cgi-bin ?
> cache_mem 100 MB
> maximum_object_size_in_memory 34 KB
> cache_replacement_policy heap LFUDA
> cache_dir ufs /tmp/ramdiskt 18000 16 256
> maximum_object_size 8 MB
> cache_swap_low 90
> cache_swap_high 95
> access_log none
> cache_store_log none
> refresh_pattern ^ftp:           1440    20%     10080
> refresh_pattern ^gopher:        1440    0%      1440
> refresh_pattern (cgi-bin|\?)    0       0%      0
> refresh_pattern .               0       20%     4320
> half_closed_clients off
> shutdown_lifetime 10 seconds
> snmp_access allow ibsnmp mgt
> snmp_access allow ibsnmp
> snmp_access deny all
> snmp_incoming_address 0.0.0.0
> snmp_outgoing_address 255.255.255.255
> icp_port 0
> log_icp_queries off
> dns_nameservers XXXXXXXX
> hosts_file /etc/hosts
> client_db off
> coredump_dir /var/spool/squid
> pipeline_prefetch on
>
> Regards.
> --
> View this message in context: http://www.nabble.com/Squid-high-bandwidth-IO-issue-%28ramdisk-SSD%29-tp24775448p24775448.html
> Sent from the Squid - Users mailing list archive at Nabble.com.
>
>
Received on Sun Aug 02 2009 - 07:14:31 MDT