(Not-so-very) Early Network Automation

If you’re not old enough to know otherwise, you’d think (based on recent hype) that we discovered network automation a few years ago. Not true. One of my readers sent me a link to excellent Managing IP Networks with Free Software presentation from NANOG26 (October 2002).

I found the presentation awesome, nothing new, and extremely sad… all at the same time.


  1. Fortunately we've come a long way since:
    1) data in (router configs)
    2) do stuff
    3) data out

    Haven't we?

    Raise your hand when the actual config in the box is not the single-source-of-truth anymore...
  2. Those of you who were involved in routing in 90ties and around 2000 should remember TeraTerm with ability to automate login process. I automated this way many tasks like collecting statistics, reconfiguring passwords...

    People involved in testing have been doing it for more than 10 years BUT it was not the hype at that time. We needed to automate repetitive tasks because we could save time.
  3. Sometimes it is really difficult to accept that people never learn. Always ignoring the past and re-discovering stuff instead of looking at the already existing solutions.
    But this is not new... Happening in the last few thousand years...
    And we still cannot product as durable houses as they did in ancient Roman Empire... The concrete they produced is still the best after 2 thousands of years...
  4. Remember when vendors had macro commands in their switches for basic automation and repeating of certain commands with variables? Recall in terminal tools such as ZOC you can record your repetitive steps, save it as a REXX macro adjust with variables is you wish and run it over and over again. I did that once years ago to deploy QoS commands 10,000 switch ports across global switches in a week for a manufacture because I could not trust the automation tool CiscoWorks NetJob.
  5. There is a difference between 20 years of experience and one year of experience repeated 20 times.
    1. This exactly. Currently we focus too much on fancy tools that we could use, instead of creating a proper solution to fix our problems.
Add comment