PFsense and Squid
-
output of requested command:
ls -l /usr/local/pkg/squid* -rwxr-xr-x 1 root wheel 49147 May 29 14:24 /usr/local/pkg/squid.inc -rw-r--r-- 1 root wheel 13976 May 29 14:22 /usr/local/pkg/squid.xml -rwxr-xr-x 1 root wheel 9454 May 29 14:24 /usr/local/pkg/squid_auth.xml -rwxr-xr-x 1 root wheel 11415 May 29 14:24 /usr/local/pkg/squid_cache.xml -rwxr-xr-x 1 root wheel 6334 May 29 14:24 /usr/local/pkg/squid_nac.xml -rwxr-xr-x 1 root wheel 11716 May 29 14:24 /usr/local/pkg/squid_ng.xml -rwxr-xr-x 1 root wheel 7310 May 29 14:24 /usr/local/pkg/squid_traffic.xml -rwxr-xr-x 1 root wheel 5047 May 29 14:24 /usr/local/pkg/squid_upstream.xml -rwxr-xr-x 1 root wheel 4362 May 29 14:24 /usr/local/pkg/squid_users.xml -
That looks fine. As long as squid.inc is there it shouldn't be complaining about any missing files. Yours is a different size than mine but that's because I checked in an unrelated fix to squid.inc for another problem this morning so it's a little bigger now.
-
That looks fine. As long as squid.inc is there it shouldn't be complaining about any missing files. Yours is a different size than mine but that's because I checked in an unrelated fix to squid.inc for another problem this morning so it's a little bigger now.
So where does this leave us re a fix?
-
If it can't be reproduced or diagnosed, it can't be fixed. I don't see any problems anywhere. I've removed it several times and reinstalled it on multiple VMs and I never see those errors. The files are all there, and yet you're seeing an error suggestion it's not actually there.
Maybe run a FS check:
fsck -p /And make sure you don't have a problem there.
-
I will run the check but unless the package installation is corrupting something on varying different hardware platforms (even virtual) then I doubt it will throw anything up.
-
I just got a write access denied and an unexpected inconsistency message. Will run fsck -f and see what happens
-
I got some unreferenced files. I will reboot and see if the problem is cured.
EDIT: I just noticed that the file references were not cleared. How do I clear them?
-
We need a good way to trigger a fsck manually. For now the best way is to boot to single user mode and run:
fsck -fy /And then reboot again.
-
I shut down after the fsck -f attempt mentioned above, and after restarting it did the check itself during startup. It automatically marked the file system clean.
I have just looked at the logs and the problem persists. The service still cannot be restarted. I have just attempted to restart the service from the console and it reports that service squid does not exist within rc.d. I am not sure if this is related.
-
How did you try to start it by hand?
It may not have an rc script if it isn't loading the .inc file properly.
Uninstall anything squid related, remove its traces:
pkg_delete -f \*squid\* rm /usr/local/pkg/squid* rm -rf /var/squid/And then reinstall squid again, see if it behaves any differently. If you had squidguard installed, or lightsquid, don't reinstall those at first, wait until squid is running.
-
EDIT: I started by running "service squid restart" from the console. Just for reference I tried from the web configuratior as well and that didn't work either,
I have just removed squid. The uninstaller told me that there was no reference to squid in the rc.d file to remove and that some folders could not be removed as they did not exist.
I have just received multiple file not found errors on bootup after a reboot as there is still a file somewhere trying to start squid.
-
I just got a write access denied and an unexpected inconsistency message. Will run fsck -f and see what happens
Maybe your hard drive has developed a bad spot and you have consequently lost a few files. You almost certainly need to reinstall the package to recover the files that have been lost. I believe modern hard drives have spare blocks that can be used to replace bad blocks. I wonder if the drive still has spare blocks. You could try reading the whole drive to see if any errors are reported, e.g. if your hard drive is /dev/ad0 the pfSense shell command
dd if=/dev/ad0 of=/dev/null bs=65536
will read the whole hard drive.
Have you recently had any panics or other restarts not preceded by a shutdown? On my pfSense system I find there are often file system troubles reported after a panic.
-
The hard drive is virtual, within VMWare. The problems are also replicated on a physical disk. Therefore this is not a hardware fault.
