SYSTEM Cited by 1 source

Salesforce¶

Salesforce is surfaced in the wiki as the customer in AWS's 2026-01-12 case study on migrating from Cluster Autoscaler to Karpenter across a 1,000+ EKS cluster / 1,180+ node-pool fleet serving thousands of internal tenants. That post treats Salesforce's Kubernetes platform as "one of the world's most complex" and the migration as a canonical production reference for Karpenter at extreme scale.

Minimum-viable page; expand on future Salesforce-infra sources.

Platform scale surfaced¶

1,000+ EKS clusters serving thousands of internal tenants.
1,180+ node pools pre-migration.
Workload mix spans mission-critical services to experimental projects.
Memory-intensive + GPU + ARM + x86 instance types in production.

Seen in¶

sources/2026-01-12-aws-salesforce-karpenter-migration-1000-eks-clusters — migration from Cluster Autoscaler / Auto Scaling groups to Karpenter across the whole EKS fleet. Canonical 1,000-cluster production reference for Karpenter.

systems/aws-eks — the runtime platform
systems/karpenter — the autoscaler Salesforce adopted
systems/cluster-autoscaler — the autoscaler Salesforce retired

Salesforce¶

Platform scale surfaced¶

Seen in¶

Related¶