Telstra outage blamed on known bug in obsolete server

Rule of thumb is, the more expensive the GPS-based thingo, the buggier its firmware. Seriously, that's from actual data, look up things like the middle- of-the-earth attack. A $5 GPS module off Aliexpress wouldn't have had this problem while a five-figure time reference can lock up or even brick itself from something as simple as getting a date set to an invalid value. I'm running an $80 device that's older than that Symmetricom and hasn't missed a beat in all that time.

Also TFA indicates a lack of understanding of how these things are run, Symmetricoms are hardware that you install once and run forever, you don't toss it just because it's a few years old like it was a mobile phone. So the "discontinued in 2016" makes for a nice sound bite for criticism but it's meaningless in this context (mine was discontinued in 2014).

Finally, following the inverse-cost-crap law for GPS things, the new one would be just as buggy as the one it's replacing even if it means you can check a box on a form that has "followed best practice" somewhere next to it.

I think the real problem, not covered in the article unfortunately, is "why did this single point of failure exist?" and "why weren't Telstra aware of the flakiness of these devices?", not specifically the rollover issue which they did know about about but the poor-quality implementation in general. It seems a little like scapegoating, "we found the cause, someone else's hardware, sorry, our bad" rather than "this whole situation shouldn't have been created in the first place".