It just means the internet is built on a very flimsy stack of technologies and any of them failing causes huge downstream issues. We saw that with AWS, and now with Cloudflare.
It’s only concerning if there are no alternatives, but as it stands there are other companies that all of these websites could have done a failover to when both AWS or Cloudflare went down. But they decided that their websites having a single point of failure was worth the risk over paying for having a proper backup system ready to go.
I like to think there was a specific person in Nebraska the author had in mind. The University there had a tap into the ARPANET back in the day and always had interesting projects going in that one wouldn’t typically expect in Nebraska.
So many people seem to have just forgotten the crowd strike outage, which halted air traffic for a day and stopped a not-insignificant amount of public infrastructure
for a solid while i had forgotten cloudflare and crowdstrike were different entities, so i spent like 5 minutes scrolling through lemmy, incredibly confused
It just means the internet is built on a very flimsy stack of technologies and any of them failing causes huge downstream issues. We saw that with AWS, and now with Cloudflare.
It’s only concerning if there are no alternatives, but as it stands there are other companies that all of these websites could have done a failover to when both AWS or Cloudflare went down. But they decided that their websites having a single point of failure was worth the risk over paying for having a proper backup system ready to go.
Relevant XKCD, as always:
XKCD//2347
I like to think there was a specific person in Nebraska the author had in mind. The University there had a tap into the ARPANET back in the day and always had interesting projects going in that one wouldn’t typically expect in Nebraska.
(Joke stolen from another post that’s since been deleted, so reproduced here.)
DNS doesn’t fail over, unfortunately.
I now imagine all the websites to fail over to the same backup services, effectively ddosing them and creating a chain reaction :D
Yeah! We call those “Cascading Failures”
They’re a nightmare! 😄
And azure also went down too
So many people seem to have just forgotten the crowd strike outage, which halted air traffic for a day and stopped a not-insignificant amount of public infrastructure
for a solid while i had forgotten cloudflare and crowdstrike were different entities, so i spent like 5 minutes scrolling through lemmy, incredibly confused