Affecting Coinbase[1] as well, which is ridiculous. Can't access the web UI at all. At their scale and importance they should be multi-region if not multi-cloud.
AWS multiple services outage in us-east-1
71–80 of 1001 posts
Re: AWS multiple services outage in us-east-1
#72Wasn't the point why AWS is so much premium that you will always get at least 6 nines if not more in availability?
You can. You just need to do the work to make it work. That's the bit where everyone but Netflix and Amazon fail.
Re: AWS multiple services outage in us-east-1
#73Even railway's status page is down (guess they use Vercel):
Re: AWS multiple services outage in us-east-1
#74Choosing us-east-1 as your primary region is good, because when you're down, everybody's down, too. You don't get this luxury with other US regions!
Sometimes we all need a tech shutdown.
Re: AWS multiple services outage in us-east-1
#75Various AI services (e.g. Perplexity) are down as well
Re: AWS multiple services outage in us-east-1
#76Meanwhile my pair of 12 year old raspberry pi's hangling my home services like DNS survive their 3rd AWS us-east-1 outage.
"But you can't do webscale uptime on your own"
Sure. I suspect even a single pi with auto-updates on has less downtime.
Re: AWS multiple services outage in us-east-1
#77Can't resolve any records for dynamodb.us-east-1.amazonaws.com However, if you desperately need to access it you can force resolve it to 3.218.182.212. Seems to work for me. DNS through HN curl -v --resolve "dynamodb.us-east-1.amazonaws.com:443:3.218.182.212" https://dynamodb.us-east-1.amazonaws.com/
Dude!! Life saver.
Re: AWS multiple services outage in us-east-1
#78Is there any data on which AWS regions are most reliable? I feel like every time I hear about an AWS outage it's in us-east-1.
Re: AWS multiple services outage in us-east-1
#79Looks like maybe a DNS issue? https://www.whatsmydns.net/#A/dynamodb.us-east-1.amazonaws.c... Resolves to nothing.
It's always DNS.
Re: AWS multiple services outage in us-east-1
#80"Never choose us-east-1"
Never choose a single point of failure.
Or rather
Ensure your single point of failure risk is appropriate for your business. I don't have full resilience for my companies AS going down, but we do have limited DR capability. Same with the loss of a major city or two.
I'm not 100% confident in a Thames Barrier flood situation, as I suspect some of our providers don't have the resilience levels we do, but we'd still be able to provide some minimal capability.