IT KORR Knowledge Center
Infrastructure Audit Checklist
A checklist for auditing physical infrastructure, network configuration, virtualization, storage, and monitoring coverage across the environment.
Physical Infrastructure
- Rack and cabling documentation is current and matches what is physically installed.
- Environmental monitoring is in place for temperature, humidity, and water/leak detection in server and network closets.
- Power redundancy is verified — UPS runtime tested, generator (if present) tested under load, and dual power feeds confirmed for critical equipment.
- Physical access to server rooms and network closets is restricted and logged.
Network Configuration
- VLAN segmentation is verified against the documented plan — no undocumented VLANs or misassigned ports.
- Firewall rule review completed — unused or overly permissive rules identified and removed or tightened.
- Unused switch ports are disabled and not left active on a default or trunk VLAN.
- Default credentials and default community strings have been changed on all network devices.
Virtualization
- Hypervisor patching is current across all hosts, with a documented patching cadence.
- VM sprawl and orphaned VM review completed — unused, powered-off, or forgotten VMs identified and decommissioned.
- Resource allocation (vCPU, memory, storage) reviewed against actual utilization to identify over- or under-provisioning.
- Host cluster redundancy (e.g., HA/failover configuration) verified functional, not just configured.
Storage
- Capacity headroom confirmed sufficient against projected growth, not just current usage.
- RAID or equivalent redundancy configuration verified on all storage arrays, with no degraded arrays left unresolved.
- Backup integration confirmed — storage systems are actually included in the backup schedule, not assumed to be.
- Storage performance (latency, IOPS) checked against workload requirements, not just capacity.
Monitoring Coverage
- All critical devices (servers, switches, routers, firewalls, storage) are confirmed present and reporting in the monitoring system.
- Alert thresholds are tuned to the environment, not left at vendor defaults that generate noise or miss real issues.
- Escalation paths are defined for each alert severity, with named on-call responsibility.
- Monitoring system itself has a health check — someone is watching the watcher.
Related Resources
- Enterprise Infrastructure Fundamentals — /knowledge-center/infrastructure/infrastructure-networking/enterprise-infrastructure-fundamentals
- Infrastructure Monitoring Fundamentals — /knowledge-center/infrastructure/infrastructure-networking/infrastructure-monitoring-fundamentals