2 ms·

Continuing the pattern of downtimes in high-profile tech products. I'm genuinely curious whether this is a quirky property of Poisson distribution or something

Continuing the pattern of downtimes in high-profile tech products. I'm genuinely curious whether this is a quirky property of Poisson distribution or something more interesting is going on... EDIT: One datapoint, Boston, MA, here still getting 500s e.g. in gist.

slondr 7y ago
The web is so centralized at this point that I don't think Poisson processes even really apply. When Cloudflare goes down, so does half the world. When AWS sneezes, 10% of websites die for an hour. It wouldn't surprise me if Azure was experiencing some funkiness and that's what was causing GitHub's issues.