squid 2.5.STABLE10 (localhost) on slackware 10.1...
i am getting an intermittent/sporadic squid DNS error when surfing the web (through squid)... it will happen with known good URLs and when it happens the "outage" only lasts less than a minute... the URL works fine if accessed directly without the proxy... the error message is usually the classic squid "
name error... the cache was unable to resolve, blah blah blah"...
the squid access log shows that there's
TCP_MISS/503 errors when the problem occurs... AFAIK everything looks as if it was an invalid URL during the outage - but once again, the URLs are known good and perfectly accessible using non-squid means at the exact same moment in which squid is giving the dns error...
as for troubleshooting, i've played with the conf file a lot, and recompiled squid several times, stripping it down a little more each time, tinkering with the configure options - the problem is not affected... BTW, i've compiled squid both with and without the internal dns function... here's the
current simplified build options i am using to compile my (still problematic) squid:
Code:
CFLAGS="-O2 -march=i486 -mcpu=i686" \
./configure --prefix=/usr \
--datadir=/usr/share/squid \
--sysconfdir=/etc/squid \
--localstatedir=/var/lib/squid \
--enable-storeio=ufs,null \
--enable-removal-policies=lru,heap \
--disable-wccp \
--enable-default-err-language=English \
--enable-poll \
--disable-ident-lookups
during the outages i can hit the refresh button in my browser repeatedly and each time squid will serve me the dns error instead of the URL... after several seconds a new refresh will successfully bring up the URL... during the outage i am able to access other URLs with squid (opened in other firefox tabs) successfully...
URLs are
seemingly being randomly affected at different times, something which has made this quite the troubleshooting nightmare for me, as i am unable to
intentionally recreate the situation (make a URL fail) on demand...
here's my squid.conf:
Code:
acl QUERY urlpath_regex cgi-bin \?
no_cache deny QUERY
acl all src 0.0.0.0/0.0.0.0
acl localhost src 127.0.0.1/255.255.255.255
acl to_localhost dst 127.0.0.0/8
acl Safe_ports port 80
http_access allow localhost
http_access deny !Safe_ports
http_access deny all
http_reply_access allow all
http_port 127.0.0.1:3128
icp_port 0
cache_effective_user squid
cache_effective_group squid
cache_dir ufs /var/lib/squid/cache 1024 32 256
cache_replacement_policy heap LFUDA
maximum_object_size 128 MB
cache_mem 8 MB
maximum_object_size_in_memory 8 KB
memory_replacement_policy heap LFUDA
cache_mgr win32sux@myisp.net
#cache_log none
cache_store_log none
#cache_access_log none
visible_hostname ellen.myisp.net
pid_filename /var/run/squid.pid
my resolv.conf (obtained via DHCP) looks like this:
Code:
nameserver 203.70.120.4
nameserver 203.70.120.5
search cpe.myisp.net
any help would be
greatly appreciated...
EDIT / UPDATE:
BTW, you can see the problem happens with wget also (when used through squid):
Squid/DNS Intermittent Issue
argh!
