[squid-users] ncsa_auth crashes squid
Summary: My squid-2.5.STABLE9-8 running on FC4 (kernel 2.6.12-1.1447) and have no problem until I introduce ncsa authentication, following the instruction found at http://www.linuxhomenetworking.com/linux-adv/squid.htm Error message WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process found in cache.log when squid starts with /etc/rc.d/init.d/squid script. Client browsers still see authentication screen, but squid crashes and restarts after the client provides userid / password and hit Enter key. In addition, many pids are assigned to squid. After crashing and restart, some more pids are added to this list. Squid starts and authenticates normally if we start squid from command line instead, i.e. type #squid. No error message found in cache.log, and client can access internet if proper id and password provided. Details: - Start squid using /etc/rc.d/init.d/squid script and see this warning in the cache.log -|- /var/log/squid/cache.log 2005/09/09 14:24:43| Starting Squid Cache version 2.5.STABLE9 for i386-redhat-linux-gnu... 2005/09/09 14:24:43| Process ID 2284 2005/09/09 14:24:43| With 1024 file descriptors available 2005/09/09 14:24:43| DNS Socket created at 0.0.0.0, port 32771, FD 5 2005/09/09 14:24:43| Adding nameserver 192.168.1.10 from squid.conf 2005/09/09 14:24:43| Adding nameserver 203.113.131.1 from squid.conf 2005/09/09 14:24:43| Adding nameserver 203.113.131.2 from squid.conf 2005/09/09 14:24:43| helperOpenServers: Starting 5 'ncsa_auth' processes 2005/09/09 14:24:43| WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process. 2005/09/09 14:24:43| WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process. 2005/09/09 14:24:43| WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process. 2005/09/09 14:24:43| WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process. 2005/09/09 14:24:43| WARNING: Cannot run '/usr/lib/squid/ncsa_auth' process. 2005/09/09 14:24:43| User-Agent logging is disabled. 2005/09/09 14:24:43| Referer logging is disabled. 2005/09/09 14:24:43| Unlinkd pipe opened on FD 10 2005/09/09 14:24:43| Swap maxSize 7372800 KB, estimated 567138 objects 2005/09/09 14:24:43| Target number of buckets: 28356 2005/09/09 14:24:43| Using 32768 Store buckets 2005/09/09 14:24:43| Max Mem size: 8192 KB 2005/09/09 14:24:43| Max Swap size: 7372800 KB 2005/09/09 14:24:43| Rebuilding storage in /var/spool/squid (DIRTY) 2005/09/09 14:24:43| Using Least Load store dir selection 2005/09/09 14:24:43| Set Current Directory to /var/spool/squid 2005/09/09 14:24:43| Loaded Icons. 2005/09/09 14:24:43| Accepting HTTP connections at 0.0.0.0, port 3128, FD 12. 2005/09/09 14:24:43| Accepting HTTP connections at 0.0.0.0, port 8080, FD 13. 2005/09/09 14:24:43| Accepting ICP messages at 0.0.0.0, port 3130, FD 14. 2005/09/09 14:24:43| WCCP Disabled. 2005/09/09 14:24:43| Ready to serve requests. 2005/09/09 14:24:44| Done reading /var/spool/squid swaplog (700 entries) 2005/09/09 14:24:44| Finished rebuilding storage from disk. 2005/09/09 14:24:44| 700 Entries scanned 2005/09/09 14:24:44| 0 Invalid entries. 2005/09/09 14:24:44| 0 With invalid flags. 2005/09/09 14:24:44| 700 Objects loaded. 2005/09/09 14:24:44| 0 Objects expired. 2005/09/09 14:24:44| 0 Objects cancelled. 2005/09/09 14:24:44| 0 Duplicate URLs purged. 2005/09/09 14:24:44| 0 Swapfile clashes avoided. 2005/09/09 14:24:44| Took 0.5 seconds (1362.7 objects/sec). 2005/09/09 14:24:44| Beginning Validation Procedure 2005/09/09 14:24:44| Completed Validation Procedure 2005/09/09 14:24:44| Validated 700 Entries 2005/09/09 14:24:44| store_swap_size = 5528k 2005/09/09 14:24:44| storeLateRelease: released 0 objects -|- Many process id are assigned for squid: [EMAIL PROTECTED] ~]# service squid status squid (pid 2122 2121 2120 2119 2118 2116 2114) is running... -|- Start client browser, authentication prompt pops, but when type in id and passwd wil make squid crash. 2005/09/09 14:28:33| WARNING: All basicauthenticator processes are busy. 2005/09/09 14:28:33| WARNING: up to 1 pending requests queued 2005/09/09 14:28:33| storeDirWriteCleanLogs: Starting... 2005/09/09 14:28:33| WARNING: Closing open FD 12 2005/09/09 14:28:33| Finished. Wrote 700 entries. 2005/09/09 14:28:33| Took 0.0 seconds (291909.9 entries/sec). FATAL: Too many queued basicauthenticator requests (1 on 0) Squid Cache (Version 2.5.STABLE9): Terminated abnormally. CPU Usage: 0.119 seconds = 0.051 user + 0.068 sys Maximum Resident Size: 0 KB Page faults with physical i/o: 2 Memory usage for squid via mallinfo(): total space in arena:2112 KB Ordinary blocks: 2052 KB 4 blks Small blocks: 0 KB 0 blks Holding blocks: 340 KB 2 blks Free Small blocks: 0 KB Free Ordinary blocks: 59 KB Total in use:2392 KB 113% Total free:59 KB 3% 2005/09/09 14:28:36| Starting Squid
[squid-users] squid wccp problem - packets not redirecting to squid
Hi all, I've a problem with redirecting wccp packets to squid. What I've done is setup gre1, loaded module ip_wccp, and I'm using mandrake 10.0, cisco IOS 12.x The problem is that the cisco router shows that the cisco has registered the server as cache-engine, and the port 80 traffic is passing through to my gre1. However, it doesn't get passed on to my squid (running at port 8080). I've already setup my iptables rule and there's no firewall blocking. iptables -A PREROUTING -p tcp -i eth1 --dport 80 -j REDIRECT --to-port 8080 (my interface is eth1) this is the tcpdump from my server (tcpdump -i gre1) ... 00:16:40.826606 xxx.xxx.xxx.xxx.3330 xxx.xxx.xxx.xxx.http: S 2830852043:2830852043(0) win 16384 mss 1460,nop,nop,sackOK (DF) [ttl 1] 00:16:40.827270 xxx.xxx.xxx.xxx.63387 xxx.xxx.xxx.xxx.http: S 3224954194:3224954194(0) win 64240 mss 1460,nop,nop,sackOK (DF) ... and this (tcpdump -i eth1 port 80) 00:17:40.277670 xxx.xxx.xxx.xxx.4158 xxx.xxx.xxx.xxx.http: S 3786887415:3786887415(0) win 16384 mss 1460,nop,nop,sackOK (DF) 00:17:40.278070 xxx.xxx.xxx.xxx.4158 xxx.xxx.xxx.xxx.http: S 3786887415:3786887415(0) win 16384 mss 1460,nop,nop,sackOK (DF) 00:17:40.278470 xxx.xxx.xxx.xxx.4158 xxx.xxx.xxx.xxx.http: S 3786887415:3786887415(0) win 16384 mss 1460,nop,nop,sackOK (DF) ... and port 2048.. 00:18:54.119886 xxx..xxx.xxx.xxx.2048 xxx.xxx.xxx.xxx.2048: [udp sum ok] udp 52 (DF) (ttl 64, id 32123, len 80) 0x 4500 0050 7d7b 4000 4011 eb92 ca9b 1e37 [EMAIL PROTECTED]@..7 0x0010 ca9b 1e21 0800 0800 003c 1c5b 0007 ...!..[ 0x0020 0004 0x0030 0x0040 0181 00:18:54.120666 xxx.xxx.xxx.xxx.2048 xxx.xxx.xxx.xxx.2048: [udp sum ok] udp 64 (ttl 255, id 46240, len 92) 0x 4500 005c b4a0 ff11 3561 ca9b 1e21 E..\..5a...! 0x0010 ca9b 1e37 0800 0800 0048 3369 0008 ...7.H3i 0x0020 0004 0004 0182 0001 0x0030 ca9b 1e37 ...7 0x0040 0x0050 but tcpdump port 8080 shows nothing... this is squid's cachelog ... Accepting WCCP messages on port 2048, FD 15. ... Any ideas? Thanks. __ Click here to donate to the Hurricane Katrina relief effort. http://store.yahoo.com/redcross-donate3/
[squid-users] Coss storage
I am using FC3 with Squid 2.5 Stable 10, I am confuse about coss storage schem for this which file system I should use like reiserfs, ext3 or ext2. thanks __ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
Re: [squid-users] Coss storage
On Fri, 2005-09-09 at 03:38 -0700, Mohammad Shakir wrote: I am using FC3 with Squid 2.5 Stable 10, I am confuse about coss storage schem for this which file system I should use like reiserfs, ext3 or ext2. Coss storage files do not touch the FS metadata once created. Ext3 is a good choice IMO (but vfat could be good too). Kinkie
[squid-users] acl dst 255.255.255.255 means no such hostname?
Hello, I found out that if I deny users going to 255.255.255.255, it is the same as if I denied going to unknown hostnames: acl bogus dst 255.255.255.255 http_access deny bogus Originally I wanted to deny customers goind to reserved IP spaces defined by RFC1918 and RFC3330: acl bogus dst 0.0.0.0/8 10.0.0.0/8 127.0.0.0/8 169.254.0.0/16 172.16.0.0/12 192.0.2.0/24 192.168.0.0/16 198.18.0.0/15 224.0.0.0/4 240.0.0.0/4 and I was very surprised that now I get access denied instead of Unable to determine IP address (aka The domain name does not exist.). Is this wanted behaviour, a side effect of something (probably when getnostname() returns -1) or a bug? Should I fill a bugreport? (squid 2.5.STABLE10 on gentoo linux) -- Matus UHLAR - fantomas, [EMAIL PROTECTED] ; http://www.fantomas.sk/ Warning: I wish NOT to receive e-mail advertising to this address. Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu. How does cat play with mouse? cat /dev/mouse
[squid-users] PATCH failure
I tried to apply patch squid-2.5.STABLE10-STORE_PENDING.patch due to the recent news of the potential DoS vulnerability but when I apply the patch I see the following message: Hunk #1 succeeded at 250 (offset -1 lines). patching file src/ftp.c Hunk #1 FAILED at 369. The other 10 Hunks of this patch succeeded. Should I be concerned that this Hunk failed ? Does this mean that the patch is bad ?
[squid-users] constant high CPU (75%) w/ Squid25-STABLE10? thought it was cache-digests, but no ...
hi all, i've squid 25 stable 10 on osx 10.4.2. it's 'front-ending' a privoxy + tor setup, with privoxy set as a cache peer w/: cache_peer 127.0.0.1 parent 8118 7 no-query default everything seems to work as hoped/expected. but, i notice my CPU util for squid is CONSTANTLY high, typically 75-95%, e.g.: 24695 squid 82.1% 1:05.52 11535 3.14M 4.90M 4.29M 40.4M even when i'm 'doing nothing' (no web browsing, etc). i'm trying to figure out why ... googling, i found that cache-digests can be a problem, so i disbaled/recompiled -- to no effect. i've turned on level 9 debug, but see nothing informative (to me, yet, anyway) after a squid restart, and while the CPU is 'hot', all i see in the logs is: 2005/09/08 23:20:28| Starting Squid Cache version 2.5.STABLE10 for powerpc-apple-darwin8.2.0... 1126246866.272 RELEASE -1 AA9FDF7558DC18F6F1925B858C3EB1FA 400 -1-1-1 unknown -1/0 GET http://127.0.0.1:8118/squid-internal-dynamic/netdb CPU Usage: 18541.909 seconds = 2274.824 user + 16267.085 sys Maximum Resident Size: 0 KB Page faults with physical i/o: 0 2005/09/09 07:21:07| Starting Squid Cache version 2.5.STABLE10 for powerpc-apple-darwin8.2.0... 1126275704.529 RELEASE -1 AA9FDF7558DC18F6F1925B858C3EB1FA 400 -1-1-1 unknown -1/0 GET http://127.0.0.1:8118/squid-internal-dynamic/netdb which leads me to believe that there's some issue with comms to/from the privoxy peer on port 8118. but, again, i'm seeing nothing in the privoxy logs ... any clues/hints as to what might be chewing up CPU, and how to fix it? cheers, richard pgppWvNZV9F0s.pgp Description: PGP signature
Re: [squid-users] strange squid
already did same result. Even system restart doesn´t solve anything. It solves by itself after a while. Regards hopping you have solutions or ideas :) --- Jose-external [EMAIL PROTECTED] escribió: you can try to clear all cache ,and restart. - Original Message - From: Daniel Navarro [EMAIL PROTECTED] To: Squid Cache squid-users@squid-cache.org Sent: Friday, September 09, 2005 8:18 AM Subject: [squid-users] strange squid Hi all, Squid is driving me crazy. Suddenly, from friday 2 squid began stopping browsing. without stopping service or error message anywhere. Just clients can´t browse through proxy but can directly to the internet. Squid restarting doesn´t solve anything. and don´t have any related warning in squid.log. It can happen even 3 times in a morning. so Any idea? Yours from Venezuela. Daniel Navarro __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/ __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/
[squid-users] disk partition locations ?
Has anyone here experimented with putting cache_dir partitions on inner/middle/outer parts of the disk, for performance reasons ? Is there really a noticeable difference in seek/total I/O performance ? -j
Re: [squid-users] PATCH failure
Dan DeLong wrote: I tried to apply patch squid-2.5.STABLE10-STORE_PENDING.patch due to the recent news of the potential DoS vulnerability but when I apply the patch I see the following message: Hunk #1 succeeded at 250 (offset -1 lines). patching file src/ftp.c Hunk #1 FAILED at 369. The other 10 Hunks of this patch succeeded. Should I be concerned that this Hunk failed ? Does this mean that the patch is bad ? Usually, in these cases you can simply look at the .rej file, see what it tried to do, then edit the appropriate file and do the changes yourself. I've done it a bunch of times, it's usually easy to spot.
Re: [squid-users] constant high CPU (75%) w/ Squid25-STABLE10? thought it was cache-digests, but no ...
hi mark, thx for the reply =) hi all, i've squid 25 stable 10 on osx 10.4.2. it's 'front-ending' a privoxy + tor setup, with privoxy set as a cache peer w/: cache_peer 127.0.0.1 parent 8118 7 no-query default everything seems to work as hoped/expected. but, i notice my CPU util for squid is CONSTANTLY high, typically 75-95%, e.g.: 24695 squid 82.1% 1:05.52 11535 3.14M 4.90M 4.29M 40.4M Perhaps , squid is in swap handling too much. Total process size seems bigger then the part in memory. Using os tools and commands , check whether squid is swapping. well, that's a point ... i'll check when back in front of the box. in the interim, i have found (stumbled across?) the (an?) apparent culprit ... changing, in squid.conf, cache_peer 127.0.0.1 parent 8118 7 default to cache_peer 127.0.0.1 parent 8118 7 no-netdb-exchange default results in my squid-related CPU util staying, now, at 1%, except, of course, when accessing a site/page ... now, as to what i've lost/broken by disabling the 'netdb-exchange' i simply do not yet know/understand (rtfm'ing ...) cheers, richard pgplampmsf13Z.pgp Description: PGP signature
Re: [squid-users] disk partition locations ?
On 09.09 09:32, john allspaw wrote: Has anyone here experimented with putting cache_dir partitions on inner/middle/outer parts of the disk, for performance reasons ? :) This is the last thing you should care about. First you should have separate drive for squid cache, then about filesystem (although some ppl blame ext2, some blame reisersfs), storage type, enough of free space (full filesystems use to have noticeable slower performance) and enough of memory for cache. Is there really a noticeable difference in seek/total I/O performance ? There are expectations that at the begin of disk there should be lower seek performance, but better performance for sequential reads, and in the middle of disk there should be higher seek performance wit lower performance for sequential reads. However ad I use two whole drives on all of my servers, I haven't checked this. -- Matus UHLAR - fantomas, [EMAIL PROTECTED] ; http://www.fantomas.sk/ Warning: I wish NOT to receive e-mail advertising to this address. Varovanie: na tuto adresu chcem NEDOSTAVAT akukolvek reklamnu postu. The 3 biggets disasters: Hiroshima 45, Tschernobyl 86, Windows 95
[squid-users] Re: PATCH failure
Pedro, Thanks for the help.. That seemed to work. Dan - Original Message - From: Pedro Timoteo [EMAIL PROTECTED] To: squid-users@squid-cache.org Sent: Friday, September 09, 2005 12:34 PM Subject: Re: [squid-users] PATCH failure Dan DeLong wrote: I tried to apply patch squid-2.5.STABLE10-STORE_PENDING.patch due to the recent news of the potential DoS vulnerability but when I apply the patch I see the following message: Hunk #1 succeeded at 250 (offset -1 lines). patching file src/ftp.c Hunk #1 FAILED at 369. The other 10 Hunks of this patch succeeded. Should I be concerned that this Hunk failed ? Does this mean that the patch is bad ? Usually, in these cases you can simply look at the .rej file, see what it tried to do, then edit the appropriate file and do the changes yourself. I've done it a bunch of times, it's usually easy to spot.
[squid-users] CUSTOM LOGIN PAGE TO SQUID (with internet policies)
I use an external program name squid_ldap_auth.(work very well) The browsers are configured with autodetect option in connection settings.(work very well) Every things work well but the auth_param basic realm clause in squid.conf file doesn't work with all kinds of browsers. REQUIREMENTS: All users must be logged to squid for internet access. Before the user login, I need to display Internet Policies. If the user login, automatically it accept the policies. Does anybody use a custom login page with squid(EX: html page)? I need: 1. a custom secure login page to squid 2. to display Internet Policies before login 3. a multiplatform solution for solaris/linux and windows users Do you have an idea if all of that are possible? If yes, what do we do? Guy Halley
Re: [squid-users] constant high CPU (75%) w/ Squid25-STABLE10? thought it was cache-digests, but no ...
On 9/9/05, OpenMacNews [EMAIL PROTECTED] wrote: hi all, i've squid 25 stable 10 on osx 10.4.2. it's 'front-ending' a privoxy + tor setup, with privoxy set as a cache peer w/: cache_peer 127.0.0.1 parent 8118 7 no-query default everything seems to work as hoped/expected. but, i notice my CPU util for squid is CONSTANTLY high, typically 75-95%, e.g.: 24695 squid 82.1% 1:05.52 11535 3.14M 4.90M 4.29M 40.4M Perhaps , squid is in swap handling too much. Total process size seems bigger then the part in memory. Using os tools and commands , check whether squid is swapping. M.
Re: [squid-users] remote 403 error through squid
Mark Elsen wrote: On 9/8/05, Thomas Reed [EMAIL PROTECTED] wrote: Having an issue with a particular website giving me a 403 forbidden error. Details: If i hit the website as http://we.bs.ite/ I get a 403 error from the webserver there, I'm guessing its denying me a directory listing. When I bypass the proxy and use the same url the page loads correctly. Interestingly, when I go through squid and enter the url as http://we.bs.ite// (note the extra slash) it comes up fine. Is this a misconfiguration on my end or the remote web servers issue? This seems to be the only site I'm having an issue with. Thanks, Thomas Could be due to , request headers not understood when squid is used. Checkout and play with : http://web-sniffer.net/ What's the difference when using squid or not (e.g.) ? You may find some clues. M. !DSPAM:4321bc59124131009210445! It seems to connect fine when i don't use squid. the web-sniffer retireved it fine as well and I didn't see anything unusal, though i'm by no means an expert on that sort of thing. This seems to be the only site I've or my users have run across problems with in the past 4 or so months of using squid. Not sure how to diagnose the issue.. Thomas
Re: [squid-users] remote 403 error through squid
On 9/8/05, Thomas Reed [EMAIL PROTECTED] wrote: Having an issue with a particular website giving me a 403 forbidden error. Details: If i hit the website as http://we.bs.ite/ I get a 403 error from the webserver there, I'm guessing its denying me a directory listing. When I bypass the proxy and use the same url the page loads correctly. Interestingly, when I go through squid and enter the url as http://we.bs.ite// (note the extra slash) it comes up fine. Is this a misconfiguration on my end or the remote web servers issue? This seems to be the only site I'm having an issue with. Thanks, Thomas Could be due to , request headers not understood when squid is used. Checkout and play with : http://web-sniffer.net/ What's the difference when using squid or not (e.g.) ? You may find some clues. M.
[squid-users] Windows Update and Microsoft Update Success!
For everyone out there looking for the way to get Windows Update and Microsoft update to work through their Squid caches, here is the recipe I found to work for me. First create a text file named windowsupdate.txt in your squid configuration directory. Place these entries in it: windowsupdate.microsoft.com update.microsoft.com activex.microsoft.com download.windowsupdate.com www.download.windowsupdate.com codecs.microsoft.com stats.updates.microsoft.com c.microsoft.com Then edit your squid.conf file. Create a new acl like: acl WindowsUpdate dstdomain -i /usr/local/etc/squid/windowsupdate.txt Prevent caching these updates with this line: no_cache deny WindowsUpdate Make sure everyone on your squid box can get to these sites with this line: http_access allow WindowsUpdate After doing these things do a squid -k reconfigure Windows Update and Microsoft Update should be working just fine now through your cache, keeping your Windows boxes up to date. Brian E. Conklin, MCP+I, MCSE Director of Information Services Mason General Hospital PO Box 1668, Shelton, WA 98584 http://www.masongeneral.com = Mason General Hospital 901 Mt. View Drive PO Box 1668 Shelton, WA 98584 http://www.masongeneral.com (360) 426-1611 = This message is intended for the sole use of the individual and entity to whom it is addressed and may contain information that is privileged, confidential and exempt from disclosure under applicable law. If you are not the addressee nor authorized to receive for the addressee, you are hereby notified that you may not use, copy, disclose or distribute to anyone this message or any information contained in the message. If you have received this message in error, please immediately notify the sender and delete the message. Thank you.
Re: [squid-users] strange squid
is it possible an old version of msn messenger crashed my squid? Regards... --- Jose-external [EMAIL PROTECTED] escribió: you can try to clear all cache ,and restart. - Original Message - From: Daniel Navarro [EMAIL PROTECTED] To: Squid Cache squid-users@squid-cache.org Sent: Friday, September 09, 2005 8:18 AM Subject: [squid-users] strange squid Hi all, Squid is driving me crazy. Suddenly, from friday 2 squid began stopping browsing. without stopping service or error message anywhere. Just clients can´t browse through proxy but can directly to the internet. Squid restarting doesn´t solve anything. and don´t have any related warning in squid.log. It can happen even 3 times in a morning. so Any idea? Yours from Venezuela. Daniel Navarro __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/ __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/
Re: [squid-users] strange squid
Ok, looking by its time I get 3 issues. 1. Excess data from... 2. likely proxy abuse 3. url too large. and clients got timeout after a long time. I see there are some patches but not seem to go to the main problem. I am on centos 4.1 and can't find any update. Hope you can help. Daniel Navarro. --- nima sadeghian [EMAIL PROTECTED] escribió: which message clients get while they are browsing. can they detect squid port? ... On 9/10/05, Daniel Navarro [EMAIL PROTECTED] wrote: is it possible an old version of msn messenger crashed my squid? Regards... --- Jose-external [EMAIL PROTECTED] escribió: you can try to clear all cache ,and restart. - Original Message - From: Daniel Navarro [EMAIL PROTECTED] To: Squid Cache squid-users@squid-cache.org Sent: Friday, September 09, 2005 8:18 AM Subject: [squid-users] strange squid Hi all, Squid is driving me crazy. Suddenly, from friday 2 squid began stopping browsing. without stopping service or error message anywhere. Just clients can´t browse through proxy but can directly to the internet. Squid restarting doesn´t solve anything. and don´t have any related warning in squid.log. It can happen even 3 times in a morning. so Any idea? Yours from Venezuela. Daniel Navarro __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/ __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/ -- Best Regards NIMA SADEGHIAN __ Correo Yahoo! Espacio para todos tus mensajes, antivirus y antispam ¡gratis! Regístrate ya - http://correo.espanol.yahoo.com/