So, starting today, I’ve begun the process of migrating my websites out of my apartment and onto the server. In another 3 hours, the data transfer should be complete and at that point I can begin testing to make sure I’ve remembered to install all the software and have the server configured correctly. Once I’m satisfied that everything is ready, I’ll update DNS to reflect the address of the new server and everyone should (hopefully) see a seamless transition as I’ll implement redirects while the new DNS settings propagate.

For those of you using Linode, do you have any suggestions for me? Nifty tips and tricks? Things to watch out for? If so, please, let me know.

UPDATE:

And the transition to Linode is complete. If you notice any problems, please let me know.

My first, and biggest problem, is that upon rebooting into Feisty, the boot process stopped for maintenance. That’s never a good sign; Linux wasn’t able to identify several of my hard drives. “That’s no big deal.” I thought. “I’ll just log in and see what’s up, I know my data is safe.” Well, it turns out that in Feisty, the way drives are mounted has changed. No longer are IDE devices mounted using “/dev/hd*#” instead, they’re assigned to “/dev/sd*#” where * is a letter that may or may not match up with the old hd* designation.

Since I have a mix of SATA and IDE drives in my machine, things got complicated pretty quickly. My /etc/fstab (which controls which drives mount where) was in shambles. The computer was trying to mount devices that had changed names. The default way this is now handled in Ubuntu is by assigning a UUID to each partition. This ID allows devices to be mounted no matter what order they’re plugged in, which is cool. If I wanted to reorder my drives and still have them mount in the same places. It’s not so cool when your drives no longer mount and you aren’t sure what’s what.

After some experimentation, I was able to identify all my drives (by mounting every device and seeing what it contained). I then found the UUIDs for each device and edited my /etc/fstab by hand. This is not what I expect from Ubuntu however. If a “normal” user (think Mom or Dad) had run into this problem, they would have been lost and probably would have never recovered their data.

A bug report has been opened and hopefully is solved soon, this is a major problem.

My second problem is that metacity no longer starts up on login. This appears to just be a setting in Gconf that needs to be fixed (things work fine with a new test account). I’ll update this post with the fix once I find it.

Update: I tracked this error down and at some point I apparently created a .gnomerc file when I was playing with compiz. It contained the following information:

export WINDOW_MANAGER=~/.gnome-compiz-manager/openbox

Why this didn’t cause a problem before, I have no idea. I deleted the file and metacity starts again.

Well, a mouse battery isn’t the only thing those notifications are good for.

I live in a pretty remote part of the country. There’s only one fiber line in and we get at least a few (5 so far this year) power outages a year. The fiber outages tend to be caused by constructions crews (mostly in Wisconsin) but wild animals, mother nature and single fault tolerance equipment make the power unreliable. A UPS is a good investment; at the very least, it will let you shut the computers down properly.

The last time the power went out, I couldn’t figure out what was beeping. That’s where the message that showed up on my screen was useful:

Ubuntu – “Hey, idiot! The obnoxious beeping is the giant battery you connected to me!”

Have a picture of an event notification you’ve gotten from Ubuntu? Post it in a comment.

That should fix the problem for you. There are a couple of bug reports that have been filed, so hopefully it gets fixed soon.

It brings up a nice interface that allows you to select software to install by category and even lets you sort them by popularity. (Ubuntu comes with a package called popularity-contest that submits back anonymous information to Ubuntu about what software is installed.)

I checked the box next to xpdf since it is one of my favorite PDF readers and pushed Apply. After confirming and asking me for my password, it began the install.

After an absolutely painless install, I head back up to my menu to find that the program launcher is already there. Gone are the days when you would have to logout/login for your menu changes to propogate.

Now, I’m not the only one that thinks this way. My friend Matthias recently posted on how Ubuntu told him he needed to charge his wireless mouse.

Jorge Castro has been talking about this for quite a while now. I’ve always considered him to be a bit of a Gnome and Ubuntu fanatic, so until recently, I’ve always taken the things he says with a grain of salt (well, I still do, but it’s a smaller grain).

So, what brought this change about? Ubuntu’s Edgy Eft release. Yeah, Dapper was polished, but it just felt like some things were either missing or just didn’t work the way I thought they should. Well, those complaints have mostly disappeared.

So, starting tomorrow, as a semi-regular feature (at least for now), I’m going to be posting some of the things that make Ubuntu a Linux desktop for me.

I tried to play one of the songs in a playlist, only to have Amarok tell me I didn’t have the codec to play .mp3s. However, what was nice to see was an option to install it.

Much to my delight, Synaptic opened up, and proceeded to install the two packages I needed to listen to my music.

Ubuntu’s 6.10 release, Edgy Eft dropped today. The official mirrors are running pretty slowly, so check out MTU LUG’s mirror for the ISOs.

I grabbed a Sarge net-install CD, dropped it in and found myself confronting the following error:

Kernel panic: VFS: Unable to mount root fs on 00:00
Press L1-A to return to the boot prom

Having found out that other people had solved this problem by putting the disks into another machine and deleting pre-existing Solaris installs, I vowed to find another way (any of you who have taken apart a U10 will understand).

I rebooted, deciding to try with a 2.4 kernel, hoping that would solve my problem. Entering linux 24 at the install prompt, I was soon greeted with the Debian Installer in all of it’s (un)glory. As much as I wanted to use Ubuntu, oh well… I’ve still got apt at least.

Over the course of this week, I’ll attempt to get everything set up to transfer my web server to the U10. Hopefully the load doesn’t eat the machine alive.

SWAT

The Samba Web Administration Tool isn’t working for some reason. Whenever I attempt to connect to localhost:901 I get an error message that says:

The connection was refused when attempting to contact localhost:901

I’ve added the necessary lines to /etc/inetd.conf and /etc/services so I’m a bit confused on this one.

/etc/inetd.conf:

swat stream tcp nowait.400 root /usr/sbin/tcpd /usr/sbin/swat

/etc/services:

swat 901/tcp #swat

Cacti

The php script that cacti runs is throwing a bunch of errors each time it runs.

PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/odbc.so’ – /usr/lib/php4/20050606/odbc.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/pgsql.so’ – /usr/lib/php4/20050606/pgsql.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/sybase_ct.so’ – /usr/lib/php4/20050606/sybase_ct.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/odbc.so’ – /usr/lib/php4/20050606/odbc.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/pgsql.so’ – /usr/lib/php4/20050606/pgsql.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning: Unknown(): Unable to load dynamic library ‘/usr/lib/php4/20050606/sybase_ct.so’ – /usr/lib/php4/20050606/sybase_ct.so: cannot open shared object file: No such file or directory in Unknown on line 0

I’m feeling pretty retarded about this… I was missing a few php4 packages and snmpd wasn’t installed.

CUPS (printing)
Whenever I try to run /etc/init.d/cupsys start it fails and throws the following in syslog:

Feb 15 23:57:41 localhost hpiod: unable to ParDevice::Open hp:/par/ANY?device=/dev/parport0: No such file or directory: io/hpiod/ppdevice.cpp 827
Feb 15 23:57:41 localhost hpiod: unable to ParDevice::Open hp:/par/ANY?device=/dev/parport1: No such file or directory: io/hpiod/ppdevice.cpp 827
Feb 15 23:57:41 localhost hpiod: unable to ParDevice::Open hp:/par/ANY?device=/dev/parport2: No such file or directory: io/hpiod/ppdevice.cpp 827
Feb 15 23:57:41 localhost hpiod: unable to ParDevice::Open hp:/par/ANY?device=/dev/parport3: No such file or directory: io/hpiod/ppdevice.cpp 827

After reinstalling cupsys, these errors are still thrown to syslog each time but printing works.

DHCP
For some reason, I wasn’t getting a DHCP lease. The ethernet adapter wasn’t coming up on boot like it should. Whenever I ran dhclient, it would throw this output:

sit0: unknown hardware address type 776
execve (/lib/dhcp3-client/call-dhclient-script, …): Permission denied
execve (/lib/dhcp3-client/call-dhclient-script, …): Permission denied
execve (/lib/dhcp3-client/call-dhclient-script, …): Permission denied
execve (/lib/dhcp3-client/call-dhclient-script, …): Permission denied
sit0 -: unknown hardware address type 776
Listening on LPF/sit0/
Sending on LPF/sit0/
Listening on LPF/eth1/00:02:b3:19:58:4c
Sending on LPF/eth1/00:02:b3:19:58:4c
Listening on LPF/eth0/00:02:b3:19:50:3a
Sending on LPF/eth0/00:02:b3:19:50:3a
Listening on LPF/lo/
Sending on LPF/lo/
Sending on Socket/fallback
receive_packet failed on sit0: Network is down
receive_packet failed on eth1: Network is down
DHCPDISCOVER on sit0 to 255.255.255.255 port 67 interval 3
send_packet: Network is down
DHCPDISCOVER on lo to 255.255.255.255 port 67 interval 5
DHCPDISCOVER on eth1 to 255.255.255.255 port 67 interval 4
send_packet: Network is down
DHCPREQUEST on eth0 to 255.255.255.255 port 67
DHCPDISCOVER on sit0 to 255.255.255.255 port 67 interval 4
send_packet: Network is down
DHCPACK from 10.0.0.2
execve (/lib/dhcp3-client/call-dhclient-script, …): Permission denied
bound to 10.0.0.50 — renewal in 38150 seconds.

There’s nothing wrong with that, except that it wouldn’t actually bind the IP address to eth0. After manually configuring the interface and setting up a route to our gateway, I had internet access. An apt-get install --reinstall dhcp3-client later, it was fixed. I’m still not sure what exactly was wrong.

Postfix
Postfix wasn’t delivering mail non-locally. It was dumping error messages in syslog that said:

Once I checked out /var/spool/postfix I was able to determine that the permissions were all wrong. Once again, apt-get install --reinstall postfix fixed the problem.

Samba

For some reason, when I attempt to connect to any shares that I have defined, it rejects my username and password. The permissions all appear correct in /etc/samba/, so I’m not sure what’s going on.

Apparently, /etc/samba/smbpasswd is not the file that Samba actually pulls passwords from. Recreating the users enabled connecting to the shares again.

WebDAV

It seems to authenticate, however throws the following errors to /var/log/apache2/errors.log:

[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] Could not DELETE /calendars/Personal.ics due to a failed precondition (e.g. locks). [500, #0]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] The locks could not be queried for verification against a possible “If:” header. [500, #0]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] Could not open the lock database. [500, #400]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] (2)No such file or directory: Could not open property database. [500, #1]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] The locks could not be queried for verification against a possible “If:” header. [500, #0]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] Could not open the lock database. [500, #400]
[Wed Feb 15 21:58:36 2006] [error] [client 10.0.10.27] (2)No such file or directory: Could not open property database. [500, #1]

WebDAV was restored to a normal state by creating the lock database /var/lock/DAV/DAVLock (chowned to www-data) as defined in /etc/apache2/apache2.conf

#LoadModule dav_fs_module modules/mod_dav_fs.so
DAVLockDB /var/lock/DAV/DAVLock