With this task, you can quickly upgrade all your Debian machines to Ruby 1.9.1 without having to go through the process each time. Just:

cap ruby:install_19

ssh [user]@[remote.server.com]

If you find yourself connecting to the same machines repeatedly, save a few keystrokes by creating a handy alias for them. Create (or edit) “~/.ssh/config” and add as many of these as your little heart desires:

Host [the alias]
HostName [domain name or IP address]
User [the account to login as]

Now you don’t have to use the full command to access the machine, just use the alias! For example, here is how I access one of my DreamHost servers:

ssh dh

The same goes for SCP! So, to secure copy a file (my_file.txt) in my current directory to the same machine I would simply issue:

scp my_file.txt dh:

Ahh… that is easy on the fingers! What else can we do with SSH config files?

 >/dev/null 2>&1

The purpose of this is to suppress any output from the command itself, because we’re not interested. I picked up this syntax years ago because it just works, but I never knew what the ’2>&1′ actually meant, until today.

The first part:

 >/dev/null

Means redirect STDOUT (the standard output stream) to /dev/null (which is basically a blackhole for bits). That’s easy enough.

The second part:

 2>&1

Means redirect STDERR (standard error stream) to the same place as STDOUT (which was just specified). STDOUT has the assigned number 1 and STDERR has the assigned number 2.

This way both STDOUT (1) and STDERR (2) are directed to /dev/null and all output of the cronned command is suppressed.

rsnapshot version:

ERROR: /usr/bin/rsync returned 12 while processing ...

rsync version:

rsync error: error in rsync protocol data stream (code 12)

It’s probably because you don’t have rsync installed on the remote host (doh!)

Let’s see how a BASH array and for loop can help clean up our FireHOL config:

You have 3 machines that need SSH access to the server. First, you can setup variable names to reference the IP addresses (or DNS names) of the machines. Put these declarations at the top of your FireHOL config for easy maintenance.

srv1="205.205.205.1"
srv2="srv2.example.com"
srv3="143.32.2.44"

Now lets see what the SSH allow declaration will look like using these variables on interface eth0:

interface eth0 public
  policy reject
  server ssh accept src $srv1
  server ssh accept src $srv2
  server ssh accept src $srv3

Notice how each host you want to allow SSH access adds another line to your configuration. This may seem trivial in my example but can add a lot of complexity as your environment grows. Is there a better way to implement? You bet.

First, lets create an array to house all of the hosts we want to provide SSH access to:

ssh_list=($srv1 $srv2 $srv3)

Next, we change the declaration on our interface to simply loop through this list of hosts and allow SSH access:

interface eth0 public
  policy reject
  for host in ${ssh_list[@]}; do
    server ssh accept $host
  done

Much better! Now we can simply add/remove hosts from our ssh_list array (at the top of the config file where all our variables are declared) and let BASH do the rest. The key here is the ${ssh_list[@]} which returns the evaluated list of hosts inside the ssh_list array. Enjoy.

Don’t take my word for it. Read it for yourself.

I was completely surprised by this release, and almost every software package in my favorite server operating system has been upgraded!

Check out the Release announcement. The list of new packages. And the Debian Administration post which includes instructions on how to upgrade

I implemented a simple fix using Ruby. Just make it start at boot by adding it to your /etc/rc.local and it should be all good.

NOTICE: automatically reboots machine

# 2>&1 redirects stderr to stdout so we can capture it
if_status = `ifconfig eth0 2>&1`
config_file = "/etc/udev/rules.d/z25_persistent-net.rules"
 
if if_status =~ /Device not found/
  config_text = Array.new
  File.open(config_file,"r") { |file| config_text = file.readlines }
  relevant_text = config_text.select { |line| line =~ /^SUBSYSTEM==/ }
  output = relevant_text.last.gsub(/ethd/,"eth0")
  File.open(config_file,"w"){ |file| file.puts output }
  system("reboot")
end

Download and use it if you’d like.

Look in the /contrib/init.d folder of your Asterisk source directory. You’ll see a file called rc.debian.asterisk. If you installed Asterisk to the default location, don’t worry about editing this file. If you installed to a different location (eg – /usr/local), change the following line in the file:

DAEMON=/usr/sbin/asterisk

Point this at your Asterisk binary. Not sure where it is? Just type which asterisk from the command line and it will show you the full path.

Next, copy the file into the /etc/init.d/ directory like so:

cp rc.debian.asterisk /etc/init.d/asterisk

(NOTE: I am renaming the file on purpose)

Now you can control Asterisk by executing this script. Make sure it starts and stops before continuing:

/etc/init.d/asterisk start
Starting Asterisk PBX: asterisk.
/etc/init.d/asterisk stop
Stopping Asterisk PBX: asterisk.

Finally, make the system run this script during the boot process:

update-rc.d asterisk defaults

Done and done. Reboot and check the process list just to be sure!

Here’s how to synchronize your Debian system’s time with network time servers, in a few simple steps (use sudo as needed):

1) Install the necessary packages:

apt-get install ntp ntpdate

2) Stop the NTP daemon for now (apt-get will start the service upon successful installation):

/etc/init.d/ntp stop

3) Manually synchronize the system clock to the NTP server pool:

ntpdate pool.ntp.org

3.5) An explanation: the NTP server daemon will fail to sync with the NTP servers if the system time is too far from the NTP servers time so its best to manually synchronize once to get your system time close enough.

4) Restart the NTP daemon:

/etc/init.d/ntp start

5) (Optionally) Verify that the NTP daemon is synchronized by checking the syslog:

grep ntpd /var/log/syslog | tail
ntpd[6348]: synchronized to 74.53.198.146, stratum 2
ntpd[6348]: kernel time sync enabled 0001
ntpd[6348]: synchronized to 128.10.19.24, stratum 1

Your output may vary, but it should look similar to mine.

That’s all for now!