RE: [squid-users] slower connections using squid (squid is slowing down all connections)

> -----Original Message-----
> From: Alex [mailto:linuxro@online.ie]
> Sent: Wednesday, September 21, 2005 10:58 PM
> To: squid-users@squid-cache.org
> Subject: Re: [squid-users] slower connections using squid (squid is
> slowing down all connections)
>
>
> Hi Chris,
>
> I changed as you suggested me:
> - cache_replacement_policy from heap LFUDA to heap GDSF.
> - announce_period 0 (this was by default too, but
> cachemanager.cgi report me
> that is ON!!!!)
> - i am not using any accel options (httpd_accel_*), all are
> set to default
> values (nothing changed here)
> - client_persistent_connections from ON to OFF
> - i increased redirect_children from 5 (default) to 50.... Am
> i wrong?!

This shouldn't make a difference, as you have not defined a redirect_program.

>
> With this new settings, i restarted squid and ... no
> improvements. Its
> slower!!!!

Ouch.

>
> At 9.40AM, I have:
> Connection information for squid:
> Number of clients accessing cache: 67
> Number of HTTP requests received: 5439
> Average HTTP requests per minute since start: 136.4
> Cache information for squid:
> Request Hit Ratios: 5min: 25.0%, 60min: 30.5%
> Byte Hit Ratios: 5min: 11.0%, 60min: 9.0%
> Request Memory Hit Ratios: 5min: 0.0%, 60min: 0.1%
> Request Disk Hit Ratios: 5min: 56.1%, 60min: 42.8%
> Storage Swap size: 3821108 KB
> Storage Mem size: 2700 KB
> Mean Object Size: 17.31 KB
> Requests given to unlinkd: 0
>
> Resource usage for squid:
> UP Time: 2393.258 seconds
> CPU Time: 21.576 seconds
> CPU Usage: 0.90%
> CPU Usage, 5 minute avg: 0.33%
> CPU Usage, 60 minute avg: 0.91%
> Process Data Segment Size via sbrk(): 29020 KB
> Maximum Resident Size: 0 KB
> Page faults with physical i/o: 0
> Memory usage for squid via mallinfo():
> Total space in arena: 29020 KB
> Ordinary blocks: 28848 KB 89 blks
> Small blocks: 0 KB 0 blks
> Holding blocks: 1784 KB 3 blks
> Free Small blocks: 0 KB
> Free Ordinary blocks: 171 KB
> Total in use: 30632 KB 99%
> Total free: 171 KB 1%
> Total size: 30804 KB
> Memory accounted for:
> Total accounted: 18153 KB
> memPoolAlloc calls: 1579234
> memPoolFree calls: 1126234
> File descriptor usage for squid:
> Maximum number of file descriptors: 1024
> Largest file desc currently in use: 162
> Number of file desc currently in use: 146
> Files queued for open: 0
> Available number of file descriptors: 878
> Reserved number of file descriptors: 100
> Store Disk files open: 0
> Internal Data Structures:
> 220836 StoreEntries
> 456 StoreEntries with MemObjects
> 438 Hot Object Cache Items
> 220761 on-disk objects
>
> Also, i have:
> Last 5 minutes:
> client_http.requests = 2.723098/sec
> client_http.hits = 0.779933/sec

Well, the box is not being stressed... What is the output of "hdparm -t /dev/sda" (I think SATA shows up as SCSI in linux)? If I set the BIOS to use the SATA drives in "Enhanced" (both SATA and PATA drives are available) or "SATA Only", I get very poor performance. Only if I set them as "Combined" (the SATA drives "act" like a PATA channel) do I get decent performance.

>
> Here comes ONLY UNCOMMENTED LINES IN squid.conf (grep -v
> ^# /etc/squid/squid.conf |grep -v ^$)
>
> http_port 3128
> hierarchy_stoplist cgi-bin ?
> acl QUERY urlpath_regex cgi-bin \?
> no_cache deny QUERY
> cache_mem 32 MB
> maximum_object_size 32768 KB
> maximum_object_size_in_memory 32 KB
> cache_replacement_policy heap GDSF
> memory_replacement_policy heap GDSF
> cache_dir diskd /var/spool/squid 20480 16 256
> ftp_user squid@mydoom.ro
> redirect_children 50
> auth_param basic children 100
> auth_param basic realm Squid proxy-caching server
> auth_param basic program /usr/lib/squid/pam_auth
> auth_param basic credentialsttl 2 hours
> auth_param basic casesensitive off
> refresh_pattern ^ftp: 1440 20% 10080
> refresh_pattern ^gopher: 1440 0% 1440
> refresh_pattern . 0 20% 4320
> acl all src 0.0.0.0/0.0.0.0
> acl manager proto cache_object
> acl localhost src 127.0.0.1/255.255.255.255
> acl lanpass proxy_auth REQUIRED
> acl to_localhost dst 127.0.0.0/8
> acl SSL_ports port 443 563
> acl Safe_ports port 80 # http
> acl Safe_ports port 21 # ftp
> acl Safe_ports port 443 563 # https, snews
> 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 localhost
> http_access allow lanpass
> http_access deny all
> http_reply_access allow all
> icp_access allow all
> cache_mgr admin@mydoom.ro
> visible_hostname proxy.mydoom.ro
> announce_period 0
> log_icp_queries off
> cachemgr_passwd *** all
> buffered_logs on
> coredump_dir /var/spool/squid
> client_persistent_connections off
>

I see nothing in here that would be causing the horrid performance that you are reporting. What do the Median Service Times (under the General Runtime Info) look like?

Chris
Received on Thu Sep 22 2005 - 12:46:47 MDT