Oracle Cloud · FinOps
OCI Right-Sizer — Compute Rightsizing with Operator Control
A Python CLI that turns OCI compute inventory and utilization telemetry into sizing recommendations and estimated cost differences. Operators review and approve changes, with live configuration checks before execution.
Stack Python / OCI SDK / OCI Compute / OCI Monitoring
Problem
Compute allocations can drift from what an application needs. Low CPU usage alone is insufficient grounds to resize: memory demand, scheduled jobs, and the financial impact all matter.
Architecture
The CLI brings OCI Compute inventory, Monitoring telemetry, and pricing into one assessment. CPU and memory are evaluated independently. Recommendations include their rationale, telemetry coverage, and estimated monthly compute cost difference, using historical billed rates where available or list pricing.
Optional OCI Generative AI explanations translate the assessment into business language. Deterministic code controls sizing, eligibility, and cost calculations.
Review and execution
scan and recommend assess workloads without changing instances. Operators
can preview a proposed resize before authorizing execution and acknowledging
the reboot. The default execution cap is one change per run.
The execution path supports Standard and Optimized Flex VMs. It checks live shape limits and baseline dimensions, then uses an ETag for a conditional update. Insufficient telemetry blocks the corresponding reduction.
Validation and limits
The workflow was tested in an OCI environment: it identified opportunities, estimated cost differences, applied selected resizes, and verified the updated configurations. Workloads were checked after the changes, as described in the technical article.
Configuration verification checks instance state, CPU, and memory; application health remains a separate check. Savings are estimates, not measured billing reductions. There is no automatic rollback, and resizing requires a maintenance window and a recovery plan.