Squid installs only partially
-
2.7.9_4 from the package manager. squid3 doesn't appear as an option for me.
-
Hmm, ok.
Guess it may need some prodding on 2.0 to work then. I have some NanoBSD VMs setup I just haven't booted them up in a while. I'll dust them off, update them, and see if I can replicate this.
-
I appreciate that. Thanks.
-
Well I just installed squid on my 2.0 nanobsd VM and it went without issue, and the settings saved fine. It shows as running under services status, and everything appears fine.
What tab were you saving settings on when you saw the error?
Perhaps there was something left over from a previous install or some other issue… Have you tried to delete the package, maybe even rm all of the squid* files in /usr/local/pkg/ and then reinstall?
-
I will try that. I was trying to save settings on pkg_edit.php?xml=squid.xml&id=0 (Proxy Server: General Settings and Proxy Server: Cache Mgmt) when I got the ro errors. In fact, when I first loaded those pages I did not see the usual default values. For example, all of the fields, such as Minimum Object Size, were blank, where I'm used to seeing default values prefilled, and the cache system was set to null.
At that point squid was running, but no web pages would load for clients that are using it. I tried unsuccessfully to deselect the interface that squid was active on, but even that failed. I had to stop the service from the Services page.
I have not installed squid on this system prior to the start of this thread. I will try your suggestion.
-
Hmm, ok.
Can you look in your config.xml and look to see if you have a "<config>" tag under <squidcache>or <squid>settings?
The defaults should come up no matter what.</squid></squidcache></config>
-
I have this, but I have already entered values in the UI and saved. The values do appear to have saved here and in the UI, despite the errors.
<squidcache> <config><harddisk_cache_size>180000</harddisk_cache_size> <harddisk_cache_system>ufs</harddisk_cache_system> <harddisk_cache_location>/cf/conf/squid</harddisk_cache_location> <memory_cache_size>2000</memory_cache_size> <minimum_object_size>0</minimum_object_size> <maximum_object_size>1000000</maximum_object_size> <level1_subdirs>4</level1_subdirs> <memory_replacement_policy>lru</memory_replacement_policy> <cache_replacement_policy>heap LFUDA</cache_replacement_policy> <cache_swap_low>90</cache_swap_low> <cache_swap_high>95</cache_swap_high> <donotcache><enable_offline></enable_offline></donotcache></config> </squidcache> <squid><config><active_interface><allow_interface>on</allow_interface> <transparent_proxy>on</transparent_proxy> <private_subnet_proxy_off><defined_ip_proxy_off><defined_ip_proxy_off_dest><log_enabled><log_dir>/cf/conf/squid</log_dir> <log_rotate>5</log_rotate> <proxy_port>3128</proxy_port> <icp_port><visible_hostname>localhost</visible_hostname> <admin_email>clarknova@msn.com</admin_email> <error_language>English</error_language> <disable_xforward><disable_via><uri_whitespace>strip</uri_whitespace> <dns_nameservers><disable_squidversion></disable_squidversion></dns_nameservers></disable_via></disable_xforward></icp_port></log_enabled></defined_ip_proxy_off_dest></defined_ip_proxy_off></private_subnet_proxy_off></active_interface></config></squid> -
Yeah that all looks fine.
-
rm all of the squid* files in /usr/local/pkg/ and then reinstall?
k, this is weird.
/etc/rc.conf_mount_rw rm -rf /usr/local/pkg/squid* rm: /usr/local/pkg/squid.inc: Read-only file system rm: /usr/local/pkg/squid.xml: Read-only file system rm: /usr/local/pkg/squid_auth.xml: Read-only file system rm: /usr/local/pkg/squid_cache.xml: Read-only file system rm: /usr/local/pkg/squid_nac.xml: Read-only file system rm: /usr/local/pkg/squid_ng.xml: Read-only file system rm: /usr/local/pkg/squid_traffic.xml: Read-only file system rm: /usr/local/pkg/squid_upstream.xml: Read-only file system rm: /usr/local/pkg/squid_users.xml: Read-only file system df Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/ufs/pfsense0 935007 185378 674828 22% / devfs 1 1 0 100% /dev /dev/md0 39406 4 36250 0% /var/tmp /dev/md1 59246 13532 40976 25% /var /dev/ufs/cf 50527 1955 44530 4% /cf devfs 1 1 0 100% /var/dhcpd/dev 192.168.255.253:/mnt/squid 205980436 131216 195385988 0% /cf/conf/squid -
run /etc/rc.conf_mount_rw first, then rm the files, then /etc/rc.conf_mount_ro when done.
-
That's exactly what I tried, but the first command appears to have no effect.
-
Ah, when I scrolled down with my mouse wheel it scrolled down that code box and I missed that.
Might have just been bad timing, if you type "mount", does it show pfsense0 as read only after running /etc/rc.conf_mount_rw ?
-
Yes.
/etc/rc.conf_mount_rw mount /dev/ufs/pfsense0 on / (ufs, local, noatime, read-only, synchronous) devfs on /dev (devfs, local) /dev/md0 on /var/tmp (ufs, local) /dev/md1 on /var (ufs, local) /dev/ufs/cf on /cf (ufs, local) devfs on /var/dhcpd/dev (devfs, local) 192.168.255.253:/mnt/squid on /cf/conf/squid (nfs) -
I can only guess then it may be your nfs mount that is causing conf_mount_rw to fail.
-
I umounted the nfs and conf_mount_rw still fails. I'll try the other slice later.
-
I booted to the other slice and things only got worse, to the point that the web UI and console wouldn't even load.
So I pulled the CF out and did a fresh reimaging.
2.0-BETA4 (i386)
built on Wed Nov 3 16:54:58 EDT 2010
FreeBSD 8.1-RELEASE-p1
Platform nanobsd (2g)I reloaded my config and installed squid. It appeared to install normally, and the Proxy server: General settings page was prefilled with the expected values. But when I changed some of those values and saved, I again got the white screen with the repeat messages of a read-only filesystem and errors on lines 69 and 70.
Still the values I entered were saved. On the Proxy server: Cache management page all the text boxes were blank, and the disk cache type was set to null. I entered the desired values for all these and saved. Again I got the error page, but the values stuck.
At this point squid appears to be functioning somewhat, although performance is low due to what appears to be excessive IO on the disk cache, which could be related, but I have no idea.
The ro errors are perplexing though. Additionally, I got the following alerts right after loading my config:
SSHD failed to start.
There were error(s) loading the rules: pfctl: DIOCADDRULE: Device busy The line in question reads [ DIOCADDRULE]:The former I have not seen before. The latter I have.
If nobody else is having these issues then I guess that points to a problem with my config, but I don't think I'm doing anything too extraordinary there, and most of my config changes were made long before I started having problems, anyway.
