Telstra outage: The night a network decided the year was 2006
netnod.se45 points by TMWNN 6 hours ago
45 points by TMWNN 6 hours ago
Human made summary: A telecom company had time synchronisation issues in a location of their infrastructure. They eventually used a single hardware GPS clock that fixed the issues but became their only authoritative clock in this location.
One day, the clock was turn off and on and it went 1024 weeks backwards because the time GPS time protocol sucks and use a week counter with too few bits.
Apparently a GPS clock can keep track of the time if it’s up and running when the week counter overflows, otherwise it has to take a wild guess. Apparently their hardware GPS clock used a hardcoded start time from its firmware instead of trusting a less reliable existing clock.
The article finishes with some AI looking suggestions to prevent such an issue to happen again.
Mine would be to have bought one or two more GPS clocks and not from the same provider.
Lots of bloviating AI-prose. Here's the original report:
https://www.telstra.com.au/exchange/what-we-ve-learned-from-...
https://www.telstra.com.au/content/dam/tcom/dynamic-media-pr...
They're pretty readable, I recommend that second one (though it leaves some questions I have unanswered). It's a good example of how latent issues with a large system can go unrecognised for a long time before a seeming unrelated change can trigger a cascading failure
One thing it doesn't make clear – the GPS card only supported the original L1 signal, which is what causes the GPS week rollover issue. The newer L2C signal has a much longer rollover period (157 years vs 19.6 years), which means no rollover until next century. If the GPS card had supported the newer L2C signal, then likely this would not have happened even if the other misconfigurations had still occurred.
> Each decision solved the problem in front of it. Nobody was asked to look at the sum of all actions.
This is the perfect description of Telstra as a company.
This reminds me of this: https://obamawhitehouse.archives.gov/blog/2015/03/26/why-we-...
I saw the talk version of that and it has lodged in my memory. The point he made was that "health.gov" was made by a bunch of teams that made sure that their part worked, but nobody was in charge of making the whole thing work.
This is sadly "normal" for large bureaucracies.
Every time I see a giant, multi-million dollar catastrophe, it's always "proper", "documented", "enterprise", "by the book", and... "a total failure". The reason is always that nobody actually cares about the final outcome, only the paperwork in front of them that they need fill out, the checkbox that needs to be ticked, or the compliance requirement that needs to be met.
This, in my opinion, is some part of how we are schooling people - do as you are told and follow the rules and some part of not having stakes in the outcome.
Direct link to the analysis report: https://www.telstra.com.au/content/dam/tcom/dynamic-media-pr...
Author needs to decide which end of the stratum stack they want as the top.
> This design from 2010 had at the top stratum 1 sources at Australia's National Measurement Institute (NMI)
> Among otherwise comparable candidates, the lower stratum carries more weight.
ai;dr
Shout out to the Jeff Geerling video about this
All I could think of while reading this was “has no one heard of ntptrace?”
It's funny, but I had the same thing happen to me once. One day, overnight, my computer decided it was 2006.
It happened twenty years ago...
;-)
Is it really so that Time Division Multiplex timing in cellular networks is based on NTP time? I stopped reading after the article implied this - it doesn't sound plausible at all.
No, it absolutely is not.
That timing requires a very tight time, frequency and phase reference. It’s derived from local high stability GNSS-disciplined time references.