Possible problem!!

Hi squiders,

There is some problem when downloding files, users get disconnect with no reason. Do we need to tune something? Are there any timeout with squid ftp? Are there any parameters for ftpget that timeout the connection? (Note that our connection to the net is slow, very slow)? Can we increase this timeout if any?

It not a question of size (I checked). I just see "abort error" in cache.log.

I'm using Squid 1.1.10 on a PPro 200MHZ + 256MB RAM, Linux 2.0.30

Squid.conf file

http_port 8082
icp_port 3134
local_ip 194.126.34.0
local_ip 194.126.43.0
hierarchy_stoplist cgi-bin ?
cache_stoplist cgi-bin ? cyberia.net.lb
cache_mem 32
cache_swap 13500
cache_swap_low 94
cache_swap_high 95
#cache_mem_low 75
#cache_mem_high 90
maximum_object_size 8000
ipcache_size 2048
cache_dir /cache1
cache_dir /cache2
cache_dir /cache3
cache_swap_log /Data/Programs/Squid/logs/metadata.log
emulate_httpd_log on
pid_filename /etc/squid.pid
debug_options ALL,1
ftpget_options -n 60 -R -W -a -A -s .gif -p http://proxy.cyberia.net.lb:8080/icons/
ftp_user squid@cyberia.net.lb
dns_children 25
request_size 300
refresh_pattern /cgi-bin/ 0 0% 0
refresh_pattern/i /welcome\.html? 3600 20% 10080
refresh_pattern/i /index\.html? 3600 20% 10080
refresh_pattern/i \.(gif|jpe?g|xbm|png|exe|zip|gz|tgz|mov|avi|wav|class)$ 10800 85% 43200
refresh_pattern/i ^ftp://.*/$ 1440 20% 5760
refresh_pattern/i ^http://.*/$ 1440 20% 5760
refresh_pattern/i ^http:// 4320 40% 43200
refresh_pattern/i ^ftp:// 4320 50% 43200

refresh_pattern . 240 40% 20160

quick_abort 30 80 100
negative_dns_ttl 3
client_lifetime 700
shutdown_lifetime 5
cache_mgr support@cyberia.net.lb
cache_effective_user squid www
visible_hostname squid.cyberia.net.lb
dns_testnames internic.net usc.edu cs.colorado.edu mit.edu yale.edu
logfile_rotate 1
append_domain .cyberia.net.lb
Received on Tue Jul 15 1997 - 08:35:17 MDT