Description

This curriculum spans the technical, operational, and organisational dimensions of cloud capacity planning, equivalent in scope to a multi-workshop operational readiness program run in parallel with a cloud migration initiative, covering measurement, modelling, automation, governance, and cross-team coordination across hybrid environments.

Module 1: Assessing Current Workloads and Baseline Performance

Decide which legacy application metrics (CPU, memory, I/O, network) to collect based on business criticality and cloud migration priority.
Implement automated data collection from on-prem monitoring tools (e.g., Nagios, Zabbix) to build historical performance baselines.
Determine the duration of performance data retention required for seasonal trend analysis versus cost of storage in the target cloud.
Classify workloads by volatility (steady-state vs. bursty) to inform autoscaling design and instance selection.
Identify dependencies between applications and databases using network flow analysis to avoid under-provisioning in interdependent systems.
Negotiate access to production environment logs with security and operations teams while complying with change control policies.

Module 2: Cloud Sizing and Instance Selection Strategies

Select between general-purpose, compute-optimized, or memory-optimized instance families based on application profiling data.
Compare vCPU-to-memory ratios across AWS EC2, Azure VMs, and GCP Compute Engine to match workload requirements.
Decide whether to use standardized instance types across environments for operational consistency or optimize per workload.
Implement right-sizing recommendations using cloud provider tools (e.g., AWS Compute Optimizer) and validate with load testing.
Balance the risk of over-provisioning (cost) versus under-provisioning (performance degradation) during initial migration.
Document instance selection rationale for audit and governance purposes, especially for regulated workloads.

Module 3: Demand Forecasting and Scalability Modeling

Apply time-series forecasting (e.g., exponential smoothing) to historical usage data to project 6- and 12-month capacity needs.
Model peak load scenarios (e.g., end-of-month processing, marketing campaigns) using stress test results and business calendars.
Integrate business growth projections from finance teams into capacity models, adjusting for product launch timelines.
Choose between predictive (forecast-based) and reactive (metric-driven) scaling strategies for different application tiers.
Validate forecast accuracy quarterly by comparing predicted vs. actual resource consumption and recalibrating models.
Define thresholds for scaling events that avoid thrashing while maintaining SLA compliance during traffic spikes.

Module 4: Designing for Elasticity and Autoscaling

Configure horizontal pod autoscalers (HPA) in Kubernetes based on custom metrics like requests per second, not just CPU.
Set cooldown periods and scaling step sizes to prevent rapid scale-in after scale-out during transient load spikes.
Implement predictive scaling using AWS Auto Scaling plans or Azure Autoscale rules with scheduled actions.
Design warm-up procedures for stateful applications to ensure new instances are fully operational before receiving traffic.
Test autoscaling behavior under failure conditions (e.g., AZ outage) to ensure capacity rebalancing works as intended.
Monitor scaling event logs to identify patterns of unnecessary scaling and refine trigger conditions.

Module 5: Storage and Data Tiering Optimization

Map application IOPS and latency requirements to cloud storage classes (e.g., gp3 vs. io2 on AWS).
Implement lifecycle policies to automatically transition infrequently accessed data to lower-cost tiers (e.g., S3 Standard-IA).
Size database volumes with growth projections, including overhead for snapshots and transaction logs.
Decide between provisioned and burstable storage based on predictable vs. variable I/O patterns.
Configure read replicas and caching layers to reduce load on primary databases and optimize storage performance.
Monitor storage utilization trends to trigger volume resizing or migration before performance bottlenecks occur.

Module 6: Cost-Aware Capacity Governance

Implement tagging policies to allocate cloud compute and storage costs to business units and projects.
Use reserved instances or savings plans for predictable, steady-state workloads after analyzing utilization history.
Set budget alerts and automated shutdown policies for non-production environments to prevent idle resource waste.
Conduct monthly cost reviews to identify underutilized resources and enforce decommissioning processes.
Negotiate enterprise discount agreements with cloud providers based on projected multi-year usage commitments.
Balance the use of spot instances for fault-tolerant workloads against the risk of instance termination during capacity shortages.

Module 7: Monitoring, Feedback Loops, and Continuous Adjustment

Deploy monitoring agents to collect real-time performance data across hybrid environments (on-prem and cloud).
Define key capacity health indicators (e.g., average CPU utilization, scaling event frequency) for executive reporting.
Integrate capacity metrics into incident management systems to correlate performance issues with resource constraints.
Establish quarterly capacity review meetings with application owners to validate forecast assumptions and adjust plans.
Automate anomaly detection using machine learning tools (e.g., Amazon CloudWatch Anomaly Detection) to flag unexpected usage.
Update capacity models based on architectural changes, such as containerization or microservices adoption.

Module 8: Cross-Functional Alignment and Change Management

Coordinate with procurement to align cloud capacity planning with fiscal budget cycles and approval workflows.
Engage application development teams early to understand upcoming feature releases that may impact resource demand.
Define escalation paths for capacity emergencies, such as unexpected traffic surges requiring immediate provisioning.
Document capacity planning decisions in architecture review boards (ARBs) to ensure consistency across projects.
Train operations teams on interpreting capacity dashboards and executing predefined scaling playbooks.
Align capacity KPIs with business outcomes (e.g., transaction throughput, user response time) to maintain stakeholder focus.