1.2.1. Lifecycle Phases and Decision Points
💡 First Principle: Each lifecycle phase has specific deliverables and decisions. The output of each phase is the input to the next—skipping steps doesn't save time; it creates debt that compounds later.
Procurement involves validating hardware against the Hardware Compatibility List (HCL)—the vendor-maintained list of approved components. Installing hardware that's not on the HCL risks driver conflicts, firmware incompatibility, and voided warranties. This is also when capacity planning happens: estimating workloads, storage growth, and redundancy requirements before purchase.
Deployment covers physical installation, OS installation, and establishing a baseline—a documented snapshot of normal system performance (CPU, memory, disk, network utilization) when the system is healthy. Without a baseline, you can't objectively determine if performance has degraded.
Operations is the longest phase: day-to-day monitoring, patch management, capacity adjustments, and responding to incidents. This is where backup jobs run, logs are reviewed, and performance is compared against the baseline.
End of Life begins when a server is no longer cost-effective to maintain or is being replaced. Critical steps: migrate data, verify no active dependencies, document the decommissioning, wipe or destroy storage media, and recycle components properly.
⚠️ Exam Trap: The HCL check happens at procurement, not deployment. Installing non-HCL components before checking is the mistake—not discovering the incompatibility during installation.
Reflection Question: Why is establishing a performance baseline during deployment so important for later troubleshooting?
