I am facing 2 problems regarding squid .
I have setup squid simply . Including allowing & blocking some urls using by url_regs. There is no dstdomain mentioned.System RAM 1.5 GB.

<1.> Some times it takes too much time to resolv/cache even google.com. Even I have used cache_dir.

<2.> Some urls/sites automatically blocking/not resolving (and prompting access denied), while I have not block such kind of url/words . I have to manually put the url in another allowing acl . But Thats too difficult , because there are so many urls. If I use http_access allow/deny all at bottom of all acls , its allowing/blocking all respectively and does not care about previous acls.

u can use dstdomain to block sites
that is too good to block site



sachin

Ok...that may works on cache/little quick browse. But I can't understand... why unnecessarily some unknown links/sites/url are automatically blocking.

Can you post the ACLs which are using?

Also, the log file contents of when the access denials occur.

Squid is simply configured

acl mynetwork src X.X.X.X
acl allowsites url_regex -i "/rule/allowsites"
acl blocksites url_regex -i "/rule/blocksites"
http_access deny blocksites
http_access allow mynetwork
http_access allow allowsites
....................................................................................

Here are some examples which are denied.But Once browsed url without squid , as these are in cache .....then no problem to browse through squid.

http://www.cisco.com/en/US/tech/tk82...800a43ee.sh180
http://www.coreims.com/?

I am not geting any reason for the problem.Also this is not good to allow this type of url one after one.

***For security reason I m unable to mention logfiles ; Very very sorry for that***

Ummm, the point was to see the regular expressions themselves, not the name of the files you have them stored in. BTW, you probably should have the mynetwork and allowsites on the same http_access line. That is, unless your intention is to let the clients at X.X.X.X use any websites they wish (unless they match the blocksites ACL) and have the allowsites ACL apply only to people not in X.X.X.X.

You run a Squid server and you don't have access to the log files? Yikes! Well, it's a shame, as it would be incredibly useful information to assist with the troubleshooting.

Yes I want users can browse all urls until match with "deny acl" . And as per organisation rule I m bound to mention any kind of system log content.

OIC. Well, surely you can at least post the regular expressions no?

If u don't mind.....can u please tell me regular expression means......HIT,MISS or some acl.

It's the stuff you wrote in the /rule/allowsites and /rule/blocksites files.

In blocksites file , I have mentioned as following:

yahoo.com
rediff.com
orkut.com
msn
porn
sex
zapak.com
hotmail.com
gmail.com
scype
hi5
sms

I didn't use dstdomain because there are few words like sms etc.
...........................................................................................
In allowsites file , I have mentioned as following: [this because of some links are denied]

http://www.cisco.com/en/US/tech/tk82...800a43ee.sh180
http://www.filecart.com/Windows/Syst...-Plus_937.html
http://www.download.com/Fashion-Stoc...-10581124.html
http://www.cisco.com/en/US/tech/tk82...800a43ee.sh180
http://www.coreims.com/?
project

Any help...please.