GCP is having major disruptions (affecting Claude, which I had no clue was built on). We built all our infra ourselves and managed within our own k8s clusters.

So we're totally unaffected! Wild what such a small team can manage all this with a little bit of elbow grease.

Probably an issue with Cloud credentials. Self managed Kubernetes is still fine for us.

what do you think about this outage? its pretty wild isn't it?

Writer. Building @pinatacloud. Working on simple static site hosting https://orbiter.host \ https://polluterofminds.com

Google GKE and apply all resources via Terraform. And a bare metal server for auxiliary activities.

Wait talk more about this. You’re running bare metal in a closet or a data center or what?

are you saying you also host LLMs in this k8s cluster you built? id love more details on that, im keen to add some GPUs to my k3s homelab but i assumed there'd be some rough edges getting all that integrated

Infra resilience is underrated until the cloud giants start sneezing and half the internet catches a cold