INFRASTRUCTURE REVIEW

Start with a conversation about your AWS environment.

Tell us what is happening and what you want to improve. The introductory scoping call is free. If a technical review would help, we’ll propose its scope, deliverables and fee before work begins.

Request an Infrastructure Review

What happens next

The introductory call

Discuss the environment, the problem and the decision you need to make. We identify relevant people and any information needed to propose the work. No technical access is required to enquire.

A commissioned technical review

If an assessment would help, the proposal confirms the systems, evidence, access, deliverables and fee. We agree those terms before investigation begins.

Implementation and handover

You can commission the recommended changes separately. The scope identifies delivery responsibilities, validation, rollback and the material your engineers need to operate the result.

Potential review deliverables

The commissioned scope can include the following. These are separate from the free introductory conversation.

  • A summary of the environment and operating constraints.
  • Prioritised findings with supporting evidence.
  • Improvement options and trade-offs.
  • Recommended implementation steps and acceptance measures.
ILLUSTRATIVE REVIEW EXTRACTSample finding · Kubernetes workload

The API’s autoscaler cannot calculate CPU utilisation.

This is a constructed example of how a finding can be documented. It contains no customer assessment, live measurements or achieved results.

Observed evidence in this example
The API Deployment has one application container with no CPU request. Its HorizontalPodAutoscaler uses CPU utilisation as a percentage of requested CPU. The configuration therefore lacks the request needed to calculate that value.
Operational consequence
CPU-based autoscaling cannot respond as intended. During a traffic increase, the service may remain at its existing replica count and requests may slow down. A load test is needed to establish the actual effect.
Recommended change
Measure the container under representative traffic. Set a CPU request from that baseline, confirm the autoscaling bounds and test the configuration in staging before a controlled production change.
Trade-offs
A higher CPU request reserves more cluster capacity and may increase node demand. A low request can trigger scaling earlier. Compare the request, replica bounds and available capacity against service performance and cost.
Acceptance measures
  • The running container has an explicit CPU request and the autoscaler reports a usable CPU utilisation metric.
  • Replicas increase under an agreed traffic profile and request latency/error rates meet the agreed service criteria.
  • Scale-down follows the configured stabilisation behaviour after demand falls.
  • The release and rollback configuration, capacity implications and service owner are documented.

Technical basis: Kubernetes Horizontal Pod Autoscaling documentation.

CLOUD & PLATFORM ENGINEERING

Tell us about your infrastructure.

Tell us what is happening and what you want to improve. The introductory scoping call is free. Any technical review follows a proposal confirming scope, deliverables and fee.

Request an Infrastructure ReviewAn introductory conversation is free. Scope and fees for any assessment or delivery are confirmed in a proposal.