Bulk IPv4 Addresses for sale, $200 a go?

IPv4 addresses have always had a value, though the price for them has always (officially at least), been $0, though ISPs often charged administration fees to people who applied for small chunks of them.

Now that IPv4 addresses have all but run out, and with the news that Microsoft has bought a large chunk of IP addresses from the bankrupt Nortel at a price of $11.25 per IP, the first question that entered my head was “How high will the price go as the addresses get rarer?”.

There’s 10,000s of organisations that each have a small chunk of IP addresses directly assigned to them, and millions with 1 or more assigned via an ISP. Each one of those organisations should now realise that the IP address has both a value and a market price, and Universities and corporations that joined the Internet early on, with their huge and under-utilised IP address ranges, will surely be looking at what they can do with them.

A follow on from this is that organisations can now put an internal cost on assigning each user an individual IPv4 address, rather than an IPv6 address, and perhaps start to justify the cost of upgrading their network. If an ISP has 1 million ADSL subscribers online all the time, that’s millions of dollars of IPv4 assets in use, and realistically not needed, that could be migrated to IPv6 then sold.

New blog location

After having various issues with tumblr over the last few months, I’ve decided to move my blog back to wordpress, hosted on my own server.

Sorry about any duplicate RSS entries you receive, if you are a subscriber you might be best removing your existing feed and re-adding it from https://ewan.im

Quick and simple version control

While I’m a big fan of version control change tracking in general, I regularly find that the main version control systems don’t do quite what I, in the role of system administrator, want.

Sometimes all I want to do is copy any individual file or set of files to somewhere permanent, and make a note of why I made that copy (e.g. making a text configuration change or updating a CSS file). Whether the file is in /etc, /usr/local/bin/ or /var/log shouldn’t matter, and whether it’s text or binary again shouldn’t make a difference.

I’ve asked and looked around, and the closest thing would be a backup script, but obviously they are more likely to mark a backup by time and date, not a reason for the backup, plus they need on-going management and installation on any server you’re using.

To come up with a solution, I’ve written a quick and simple Bash shell script (which should work on AIX, Solaris, etc, with Bash installed), which:

  • Asks for the reason for the copy
  • Copies all files specifed on the command line to /var/echange/ with a time and date appended to the name
  • Creates a text file named after the new copy of the file with .echange added which contains the original name and location of the file, and the reason for the change.

To make a copy I just run

ec index.html test.css

And I’ll be asked for my reason, simply end the input by pressing return, and the files will be copied.

If I want to find an old copy of the file, this way I can just run

find /var/echange -name “originalfilename.*”

and find all versions of the original file that I’ve copied, and the reason why I made the copy.

I don’t know if anyone else will find this of use, but since I want it to be available for me on various systems, I’ve uploaded the file to http://www.ewanleith.com/files/ec.txt – Simply download the file to your Linux server, copy it to /usr/local/bin/ec and mark it as executable, then create /var/echange and you’re good to go.

It can also work with any other directory than /var/echange/, simple change the line at the top of the bash script to wherever you want.