Hacker News Reader

why wouldn't they?

Because it's terrible and has always been terrible?

After 2017 I'll never use US-east-1 again. Hell... I should have learned that particular lesson in 2011 but it took two catastrophic failures for me to figure it out.

There are numerous threads here on HN covering the topic "why does US-east-1 suck so hard."

https://news.ycombinator.com/item?id=13756082 is just one example.

Maybe it’s terrible because they host their own stuff there.

That's the story of AWS as a whole.

I wonder if there’s too much moral hazard for them to run a data center just for Amazon but run exactly the same as the other data centers? Or would Amazon scope creep it into its own snowflake.

I kind of doubt that they’re getting efficient use of hardware Amazon could be using but isn’t, since if someone allocates it it’s not available for Amazon anymore.

If their stuff was all super resilient and multi AZ, it would be easier to place blame on the customer

Multi AZ != Multi region. AWS has many services that run in multiple regions but some of their services are not.

Correct, I mean region. Thank you

Reply on news.ycombinator.com