Can't Change Squid Cache Size
-
Hi all,
I'm running Squid3 on the latest PF Sense. I've been pretty much running it with defaults and now I am trying to fine tune it a bit more. Tried to adjust the disk cache size from 100mb to 2048. Squid fails to start after restarting the service. I do not see any errors generated on the system log page.
Jun 6 08:43:29 squid[8837]: Squid Parent: child process 9169 started Jun 6 08:40:38 dansguardian[53333]: Error connecting to proxy Jun 6 08:38:57 squid[49714]: Squid Parent: child process 50228 started Jun 6 08:38:46 squid[40249]: Squid Parent: child process 40832 started Jun 6 08:38:44 squid[58522]: Exiting due to unexpected forced shutdown Jun 6 08:38:44 squid[58522]: Squid Parent: child process 58895 exited due to signal 15 with status 0 Jun 6 08:38:29 check_reload_status: Reloading filter Jun 6 08:38:29 php: /pkg_edit.php: Reloading Squid for configuration sync Jun 6 08:38:29 php: /pkg_edit.php: [squid] xmlrpc sync is starting. Jun 6 08:38:27 check_reload_status: Syncing firewall Jun 6 08:37:57 check_reload_status: Reloading filter Jun 6 08:37:57 squid[58522]: Squid Parent: child process 58895 started Jun 6 08:37:57 php: /pkg_edit.php: Starting Squid Jun 6 08:37:57 php: /pkg_edit.php: [squid] xmlrpc sync is starting. Jun 6 08:37:55 check_reload_status: Syncing firewall
Does anyone have any thoughts?
Thanks!
-
try to run squid -z on console/ssh
-
Thanks for your help.
Tried the command and it created the directories, however Squid still does not load.
Looking in the Squid config file I see my settings are reflected.
acl dynamic urlpath_regex cgi-bin \? cache deny dynamic cache_mem 200 MB maximum_object_size_in_memory 32 KB memory_replacement_policy heap GDSF cache_replacement_policy heap LFUDA cache_dir ufs /var/squid/cache 2048 16 256 minimum_object_size 0 KB maximum_object_size 2048 KB offline_mode offcache_swap_low 90 cache_swap_high 95
However running Squid from the command line does start Squid and is reflected as so in the services page.
If I restart Squid it fails to load again. So I take it Squid will fail to start if I reboot also?
Thoughts?
Thanks!
-
Try to start it on console to get the error.
squid -NsXY or /usr/local/etc/rc.d/squid.sh start.
-
Thanks for your reply.
Running: /usr/local/etc/rc.d/squid.sh start I saw the following:
2012/06/07 09:14:44| WARNING: (B) '127.0.0.1' is a subnetwork of (A) '127.0.0.1' 2012/06/07 09:14:44| WARNING: because of this '127.0.0.1' is ignored to keep splay tree searching predictable 2012/06/07 09:14:44| WARNING: You should probably remove '127.0.0.1' from the ACL named 'ext_manager'
So I removed those (though I had not set it, I believe that is from Dansguardian)
Does not solve my issue, though the error is now gone. As mentioned previously Squid started via command line fine. Its just when using the web portal that Squid fails to start. I tried putting the options back to the default (100M and 8k max size) and it fails to load now also… these were the only options I had altered. I'm starting to wonder if this is just an issue with the web interface and not the configuration...
Managed to reboot the firewall and Squid starts normally with my changes intact. (the rules I removed above 127.0.0.1..etc have been put back automatically by the system also)
Thoughts?
Thanks
-
As mentioned previously Squid started via command line fine. Its just when using the web portal that Squid fails to start.
This is fixed on 2.0.2 and on 2.1, apply this patch on /etc/inc/service-utils.inc and see if you can restart services on pfsense.
https://github.com/bsdperimeter/pfsense/commit/9a1248b3a813ccfad927362e8185879cdba26925
att,
Marcello Coutinho -
Thanks for the info! I copied and pasted the info over the existing file (not sure if there is a better method to for this) and it appears to be working now. Tried restarting the service = worked. Tried stopping and then starting = works.
I believe that did it. :)
Cheers!