Hacker News Reader

Good question. At 100x, probably not. At 10x, yes would be better than managing services on our own. By that time, we would have a better prioritized list of which services to self-manage and which ones to leave to AWS. Are you specifically concerned about DynamoDB for some reason?

View on HN

How easy or hard would it be to switch to self-managed components as you grow from 10X to 100X? Quite often, they end up becoming a tech debt that remains in the back burner. Just curious.

Ah yes. The engineer would tell you we can move when we want. The manager would tell you it is harder than it looks. Management would tell you it will never happen. :-)

See it as reducing startup risk and deferring the payment to when you become successful and have money/time to throw at problem. Though there are best practices to do it in a clean way so moving is easier.

Do you know some known gotchas here?

I’d be curious why you think at 100x is where you would lose out on TCO with self managed. I feel like staff time commitment should only go up with larger fleets, you’d really start running into the pricing advantages on 0-rated network here etc.

Reply on news.ycombinator.com