Analyzing and optimizing technical and business processes
Analyzing and optimizing technical and business processes
This domain tests optimizing cost, performance, and operational processes: using monitoring data, right-sizing, cost controls, and improving business continuity.
Core rules
- Cost optimization: committed use discounts (CUDs) for steady workloads; sustained use discounts automatic; Spot VMs for batch; autoscaling; Cloud Storage lifecycle policies; BigQuery partitioning/clustering; billing export to BigQuery; budgets and alerts.
- Performance: use Cloud Monitoring metrics, Cloud Trace for latency, Cloud Profiler for CPU/heap; right-size VMs from utilization; managed services (autoscaling, load balancing); Cloud CDN for static content; choose regions near users.
- Process optimization: automate with IaC and CI/CD; Cloud Deploy for release management; Error Reporting and Logging for diagnosis; define SLOs and error budgets to balance reliability and velocity.
- Business continuity: define RTO/RPO; backups (Cloud SQL automated backups, GKE Backup); multi-region replication (Cloud Storage dual-region, Spanner multi-region); disaster recovery drills.
Worked example
A company’s BigQuery costs are rising. Reason: analyze the billing export; identify expensive queries; add partitioning and clustering to reduce scanned data; set custom quotas; move steady BI workloads to flat-rate slots; apply lifecycle policies on Cloud Storage. This reduces cost without changing SLAs.
Traps
Over-provisioning to avoid performance issues; ignoring sustained use discounts; not using labels for cost allocation; missing SLOs and error-budget alerts; using Spot VMs for stateful workloads; optimizing cost without measuring impact on SLOs.