Hacker News Reader

They did have a major outage for a few minutes on 2013: https://www.theregister.co.uk/2013/08/17/google_outage/

So this is now the longest outage ever, right?

It's completely different mode of failure - when part of search is down somewhere, it's just routed to the same service in other location. If part of the GCP is down anywhere, some customers are affected.

Since original Google infrastructure was developed specifically for first kind of services, cloud org still has problems adopting it to its needs.

Sounds like a good design — but in the early days of google cloud — 2008 — you’d get the google homepage on a bad request

AFAIK google.com is running independently on a different infrastructure, but some of other services are relying on GCP hence the problem affects them too.