Thanks Layman - I have used the AT&T failure many times to underline the need for a robust approach to software changes.
I have spent most of the past 40 years undertaking commissioning and return to operational service tests on aviation systems throughout the world. As the Optus failure unfolded last week I couldn’t help but think, “Have they not understood the past experiences and applied them?” But then I remembered the issues and arguments I have frequently had recently with the “new generation” of engineers and software and network “specialists” and the answer was immediately apparent, as was the thought, “We’ll be seeing much more of this unless they are forced to apply more rigorous procedures and test regimes.”
Hopefully not only Optus but all other software dependent system providers will learn from last week’s experience and more robustly apply measures such as regression testing and failsafe step back steps.
And there are important lessons for the general public, particularly those who rely on external networks for their business and social support: Have some form of non dependent redundancy and demand a service level agreement. The former will give you peace of mind in daily operations and the latter will also give you peace of mind in normal operations and ensure, in the worst case, you have some recourse to compensation.
The fact this thread has not elicited any direct aviation issues from the Optus event hopefully means lessons have been learnt and applied.
Gne