Debian Time Machine Server

As usual, a little background for perspective. This the second time I have had to do this and to get it working I had to go scouring the Internet a little bit. This involved taking bits and pieces from a few sites then keeping my fingers crossed that it would work in the end.

There are two Macs in my household. I configured this at first so that both would dump their backups into one folder. This worked well but I noticed a few weird stuff from time to time and I remembered also reading were some people suggested keeping them separated because of issues that, I guess, could cause the stuff I was noticing. Because of this an a few other cosmetic factors, I decided to move the service to a different server and assign both computers to separate folders (I used separate LVM volumes so that they will have no effect on each other and I have flexibility to grow)

Installing Packages

The newer releases of OS X requires Netatalk 2.2.x+. However, Debian 6.0 (Squeeze) comes with 2.1, which won’t work with Mac OS X 10.8 “Mountain Lion”. If you are still running Debian 6.0 you can get netatalk 2.2 from Debian 7.0 (Wheezy) by doing the following as root.
First add the following line to /etc/apt/sources.list:

deb http://http.debian.net/debian wheezy main contrib non-free

Then run the following commands:

 aptitude update
 aptitude install netatalk avahi-daemon avahi-utils

You can revert the changes to /etc/apt/sources.list now and run “aptitude update” again. Obviously if you were already on Wheezy you won’t have to worry about this.

Setting up Netatalk
Let’s do some configs…

Change your /etc/netatalk/AppleVolumes.default file to export the Time Machine volume.

Look for the following line:

 #:DEFAULT: options:upriv,usedots 

And change it to something like this.  Also remove the hash sign:

 #:DEFAULT: cnidscheme:dbd options:upriv,usedots 

At the end of the file you’ll find a line that reads:

~/                     "Home Directory"

Add something like this below it:

/mnt/timemachine  "Time Machine"  allow:username cnidscheme:dbd volsizelimit:250000 options:usedots,upriv,tm
 

• /mnt/timemachine is your backup folder.
• “Time Machine” is a random label to identify your Time Machine volume.

The rest of the line contains various parameters to allow the Mac to “play nice” with this server as a Time Machine target. It’s important to add the options:tm at the end of the line so that Netatalk enables various special options for Time Machine. You can also add fancy options to restrict access to users logging in with specified accounts. But I have decided to keep it simple, at least for this round 😉

The next config file is /etc/netatalk/afpd.conf. Comment the last line like this:

# - -tcp -noddp -uamlist uams_dhx.so,uams_dhx2.so -nosavepassword

…and add this:

- -tcp -noddp -uamlist uams_dhx.so,uams_dhx2_passwd.so -nosavepassword

I guess you could also just replace it but I like an easy rollback path, just in case.

I am not sure if this command is actually needed for it to work but it didn’t cause it not work 🙂

touch /mnt/timemachine/.com.apple.timemachine.supported

Restart netatalk for the new configuration to take effect:

sudo service netatalk restart

For an additional layer of security I decided to create a dedicated user account that will only have access to the write to the backup folder. Time Machine will ask for this information on initial setup.

 sudo useradd -s /bin/false timemachine
 sudo passwd timemachine
 sudo chown -R timemachine:timemachine /mnt/timemachine
 

This takes care of the server side.

Client Setup
Now configure your OS X installation so it sees unsigned time machine volumes. Open the terminal app and execute the following command:

defaults write com.apple.systempreferences TMShowUnsupportedNetworkVolumes 1

PS.
Older articles refer to creating a /etc/avahi/services/afpd.service file. With netatalk 2.2, this file is redundant: you do not need to create it.

RaspBerry Pi – 1st Encounter

Back Story

I have wanted one of these boards to replace my Apple TV2 running XBMC for a long while now but as everybody knows the first version sold out almost immediately and the average price for one on the after-market is sometimes 3-4 times what the Element14 sells it for. They announced the second version and I decided to try get my hand on one before it hit the after-market. Long story short, after about a month of waiting on order fulfillment, I have it.

What to do?

At my office a couple of the guys have these configured as HTPCs (XBMC of course). I couple have positive feedback while others are so-so. The number one complaint I have heard it the speed. and digging further, I find a couple of them attribute it to slow SD cards. So I was worried that I might have to purchase another SD even though I already have one laying around. Of course I decided to try what I have on hand first, and I am happy with the results so far.

Here is what I have:

Hardware

• Raspberry Pi Model B 512MB
• Transcend 8GB Class 6 SDHC

Software

• Raspbmc

 Configuration

• Installed with python script (install.py) from the site with a downloaded image
• Install NTP
• sudo dpkg-reconfigure tzdata for America/Montreal
• Inside XBMC (Settings > Appearance > International) set location to Canada/Montreal
• I use NFS shares, so I add those with the appropriate scrapers and allow it to scan the library

Conclusion

This is not my final setup but I think this gets me back to where I was before and I am okay with that. The main thing I wanted to test here is the performance of the SD card with RaspBMC and that seem to be working out. I don’t know if Class 4 would be too slow or if the doubling of the RAM from the previous model is making all the difference here (512MB vs 256MB). I don’t have a class 4 to test the difference, so all I know is this setup works…..for now 😀

Backup PostgreSQL Schemas

I have been handed the responsibility to manage the backups of a PostgreSQL database. Easy enough you would say, but there is a catch.

1. The owner wants backups to be kept of each individual schema within the database
2. I have never managed a PostgreSQL DB before

So not being one to back down from such a challenge, I searched high and low all over the Internet, but either the Internet community isn’t interested in doing this or I am not looking in the right places. So here, with the help of fragments of scripts from around the web, I have worked out a solution which accomplishes task at hand. It is all bottled in the script below which I think is self-explanatory with the aid of a generous helping comments scattered throughout the file.

#Description:
#This script will create a compressed backup of the Genesis Postgres Db and store it on a predefined folder.
#Backups that are older than 30days will also be removed automatically

####### Make and change to directory where backups will be saved #######
BASE_DIR="/path/to/backup/folder"
YMD=$(date "+%Y-%m-%d")
DIR="$BASE_DIR/$YMD"
mkdir -p $DIR
cd $DIR

####### Full Postgres Backup #######
sudo -u postgres pg_dumpall | gzip -c > All_Db.sql.gz

####### Individual Schema Backups #######
# 1. Select individual schemas within the database and pipe the results into sed which does Step 2
# 2. Clean up the output from the SQL above
#     - Get rid of the empty spaces at the beginning of each line
#     - Remove the head and tail info from the file(Title, labels, etc)
for schema in $(sudo -u dbOwnerUN psql -d DBName -c "SELECT schemata.schema_name FROM information_schema.schemata;"|sed 's/^[ \t]*//;s/[ \t]*$//;1,2d;N;$!P;$!D;$d');
do
sudo -u dbOwnerUN pg_dump -Ft -n "$schema" DBName | gzip -c > "$schema".sql.gz
done

####### Delete backup files older than 30 days #######
OLD=$(find $BASE_DIR -type d -mtime +30)
if [ -n "$OLD" ] ; then
echo deleting old backup files: $OLD
echo $OLD | xargs rm -rfv
fi

As I said, this is my first attempt at anything like this so it may not be the best or easiest way of accomplishing this. (I am just sharing what I know and recording for my reference). So please post comments if you have a better solution.

….Questions are also welcomed

How to Tell What Version of Ubuntu You Are Running

Telling what version of Ubuntu you are running is extremely easy. You would commonly use this command to figure out if you are running Edgy after you upgraded from Dapper.

cat /etc/issue

Ubuntu edgy (development branch)

Note that the version numbers might change over time. I’m running the beta version so that’s what shows up when I run that command. Either way, it should be clear that you are running Edgy.

via How to Tell What Version of Ubuntu You Are Running – How-To Geek.

Linux on the iPhone: iPhone 3G binaries!

I wrote up a how-to for PC World on how to put Android on the iPhone 3G and the iPhone 2G and it went up today. I wanted to be there to tweet about it when it went up, but I’ve been keeping really strange hours lately and I wasn’t awake for it when it went up.

But here are the binaries (for iPhone 3G, and for iPhone 2G), graciously hosted by PC World!

Please read the how-to that I wrote for PC World to on installing it. The steps are basically the same as before, except you can put the firmware in a directory on the iPhone OS data partition. This means that you don’t have to modify the ext2 partition as before.

One thing I didn’t mention is that you could perform the installation on OS X without a Linux VM if you recompile loadibec and oibc. Otherwise, the directions are the same.

via Linux on the iPhone: iPhone 3G binaries!.