> One crucial way in which we do this is through the democratization of highly curated data sources that sunshine usage and cost patterns across Netflix’s services and teams.
No engineer has ever said this. I miss when engineering blogs were actual engineering. This blog post has been through so much management waffle rewording it no longer adds much to the conversation.
Once you hit a certain scale anything on the AWS console becomes impossible to use. It’s one of the reasons most high to very high scale companies use something else to manage cloud resources. See: Spinnaker and Asgard for Netflix specific examples, but also Terraform and the many alternatives.
I work for an organisation that is at this scale on AWS. Of course, we use Terraform, but the cost explorer and tags and so on are still usable (and useful).
For <business reasons> we have in the order of 10^6 SNS topics in one account+region pair. In that account, while in that region, the SNS console is entirely frozen. I suspect they're doing some looped pagination polling in the background and fall over on our degenerate use case.
Edit: I can also back up the other parent comment - tagging still works fine (and I know for a fact the stability and scalability of the tag system is taken more seriously than most things within AWS)
No engineer has ever said this. I miss when engineering blogs were actual engineering. This blog post has been through so much management waffle rewording it no longer adds much to the conversation.