3 ms·

You know, it would be cool if you found stats on the downtime metrics of these various high-profile recent outages, and calculated the odds of having such a clu

You know, it would be cool if you found stats on the downtime metrics of these various high-profile recent outages, and calculated the odds of having such a cluster. Statistics is hard, though, and avoiding a "Texas Bulls-eye" would be hard.

wtallis 7y ago
"Celebrities die 2.7183 at a time": http://ssp.impulsetrain.com/celebrities.html http://ssp.impulsetrain.com/celebrities.html
- gcbw2 7y ago
  So the only take away is that now the population at large notices tech companies outages as much as they notice celebrity deaths?
  - uxp 7y ago
    "population at large" This thread is linked to a status page run by Twitter, on a programming and technology news site. I'm not really seeing how most people that exist in the western/1st world are noticing this. Is there a CNN article, or FoxNews segment on how tech companies are having outages?
    - gcbw2 7y ago
      Yes, fox news even suggest it was part of a large coordinated censorship effort on the POTUS :D https://www.foxnews.com/tech/twitter-suffers-widespread-outage https://www.foxnews.com/tech/twitter-suffers-widespread-outa... quote from that url: "The outage came as President Trump was hosting a social media summit with right-wing personalities and tech industry critics who've accused Twitter and other websites of having an anti-conservative bias."
appleiigs 7y ago
Already did that here: https://news.ycombinator.com/item?id=20356610 https://news.ycombinator.com/item?id=20356610
- rossdavidh 7y ago
  Sure does look like we are way out there at the tail end of the probability distribution, by those numbers.
mruts 7y ago
I mean, we can assume the downtime variance follows a normal distribution. It should pretty easy to calculate P<.05 with just a little bit of data.
- rossdavidh 7y ago
  What you say could be true, but I don't know that we can assume it. If downtime requires several things to happen (cascading errors), but those things interact somehow (problem with one makes another more likely), I could imagine it might not be normally distributed. Disclaimer: I Am Not A Statistician.

wtallis 7y ago

"Celebrities die 2.7183 at a time": http://ssp.impulsetrain.com/celebrities.html http://ssp.impulsetrain.com/celebrities.html

gcbw2 7y ago
So the only take away is that now the population at large notices tech companies outages as much as they notice celebrity deaths?
- uxp 7y ago
  "population at large" This thread is linked to a status page run by Twitter, on a programming and technology news site. I'm not really seeing how most people that exist in the western/1st world are noticing this. Is there a CNN article, or FoxNews segment on how tech companies are having outages?
  - gcbw2 7y ago
    Yes, fox news even suggest it was part of a large coordinated censorship effort on the POTUS :D https://www.foxnews.com/tech/twitter-suffers-widespread-outage https://www.foxnews.com/tech/twitter-suffers-widespread-outa... quote from that url: "The outage came as President Trump was hosting a social media summit with right-wing personalities and tech industry critics who've accused Twitter and other websites of having an anti-conservative bias."

appleiigs 7y ago

Already did that here: https://news.ycombinator.com/item?id=20356610 https://news.ycombinator.com/item?id=20356610

rossdavidh 7y ago
Sure does look like we are way out there at the tail end of the probability distribution, by those numbers.

mruts 7y ago

I mean, we can assume the downtime variance follows a normal distribution. It should pretty easy to calculate P<.05 with just a little bit of data.

rossdavidh 7y ago
What you say could be true, but I don't know that we can assume it. If downtime requires several things to happen (cascading errors), but those things interact somehow (problem with one makes another more likely), I could imagine it might not be normally distributed. Disclaimer: I Am Not A Statistician.