and also did you guys discussed this <https://www....
# office-hours
I was very skeptical at reading it but there is some truth to a lot of it
@Erik Osterman (Cloud Posse) @Matt Gowie ( since you tried crossplane)
e
I can't recall, but I know I replied on LinkedIn
I agree, it's also why we support publishing to deployment repos in atmos. E.g. even if you use FluxCD, ArgoCD, Crossplane, you still need a process for that.
Also, I think the article is fair:
I see so many of these things slamming Terraform, but stick everything in Kubernetes; but I don't think it's as simple as that.
p
yes, and the TLDR of the article point out that TF is really good at landing zone and slow changing infra etc
but not for the fast changing app infra on top of the foundational resources etc
e
I also think there's another side to this that I remain skeptical about; not that it can't be done, but that with my strongly held beliefs won't be done well. It's one of the things we emphasize in atmos and atmos pro. When your services reach into the hundreds, thousands, tens of thousands, understanding what is deployed is impossible
In atmos pro, this is central to the whole product. So you get cross repo visibility in a centralized control plane.
What services are flapping/permadrift? What are are the deprecation warnings? what are the terraform failures? and looking at that across regions?
I'm much more bullish about things like KRO, ACK + ArgoCD/Flux than I am about crossplane
@Matt Gowie @PePe Amengual et al, it would be a great discussion
👍🏽 1