Network Operations
Operations is the discipline of running a network you can trust on its worst day
Designing and building a network gets it working once; Network Operations keeps it working, and keeps it recoverable, every day after. The whole domain hangs on one idea: a network you can run safely is a network you have written down, change deliberately, watch continuously, and can rebuild from a known-good copy. Each subtopic owns one of those activities. The classic exam trap here is treating an operational decision as a technical one. When a question asks how a firmware upgrade reaches production, the answer is not a CLI command, it is the change-management process; when it asks how you would restore service after a fire, the answer is not a faster router, it is the disaster-recovery plan and its RPO (recovery point objective) and RTO (recovery time objective) targets. Read the verb in the question: "document", "approve", "monitor", "recover", and "restore" each point at a different subtopic of this domain.
The domain unfolds as a run-the-network loop in eight steps
Walk the subtopics in order and they form the operational lifecycle. Network Documentation comes first because you cannot operate what you have not recorded: diagrams, IPAM (IP address management), SLAs (service-level agreements), and baselines are the system of record everything else reads from. Life-Cycle Management tracks each asset from purchase to retirement, so end-of-support gear is replaced on a refresh cycle before it becomes an unpatchable risk. Change Management forces every production change through one request-assess-approve-schedule-implement-review path, classified by risk. Configuration Management is where those changes land: it keeps the running, startup, and backup configs straight, holds a golden baseline, and version-controls every revision. Network Monitoring then watches the running network through SNMP (Simple Network Management Protocol), flow data, and centralized syslog, alerting on deviation from baseline. Disaster Recovery plans for the day operations fails anyway, sizing RPO and RTO and pre-staging recovery sites. Network Services covers the always-on infrastructure, DHCP for addressing, DNS for naming, and NTP for time, that the rest of the network silently depends on. Remote Access and Management closes the loop with how you reach devices to do all of the above: in-band or out-of-band, secured by SSH or a VPN.
When two answers both work, the exam rewards the documented, reversible, defense-in-depth one
Across this whole domain the safe instinct is the same: prefer the choice that is written down, can be undone, and degrades gracefully. A change with a tested back-out plan beats a faster change with none. SNMPv3 beats v1 or v2c because the older versions send credentials in cleartext. A restore from the most recent backup config beats rebuilding settings from memory. Out-of-band management beats in-band when the network itself might be the thing that is broken. When a question offers a convenient option and a disciplined one, the disciplined one is almost always keyed, because operations is judged by what survives the bad day, not by what is quickest on a good one.
The run-the-network loop: which subtopic answers which operational question
| Activity | The operational question it answers | Drill into |
|---|---|---|
| Record | Where is everything, and what does normal look like? | Network Documentation |
| Plan the asset's life | When does this device stop being safe to run? | Life-Cycle Management |
| Govern change | How does any change reach production safely? | Change Management |
| Control the config | What is each device set to, and how do I get it back? | Configuration Management |
| Watch | Is the running network healthy, and what changed? | Network Monitoring |
| Survive a disaster | How much can we lose, and how fast must we be back? | Disaster Recovery |
| Deliver core services | What addressing, naming, and time keep the network usable? | Network Services |
| Reach the devices | Which path do I manage over, and what secures it? | Remote Access and Management |