High Availability, Disaster Recovery and Network Documentation (Network+)
What you’ll know
How to measure availability and recovery, how networks avoid single points of failure, how to back up and test, and which documents keep a network fixable at 3 a.m. when the person who built it is on holiday.
Why it matters. Network Operations is almost a fifth of the exam, and this is the part that saves careers: the outage that lasts two minutes instead of two days because someone planned and wrote things down.
The four numbers
| Term | Means | Example |
|---|---|---|
| RTO (recovery time objective) | How long you can be down | “Email back within 4 hours” |
| RPO (recovery point objective) | How much data you can lose, measured in time | “At most 1 hour of orders” = back up at least hourly |
| MTBF (mean time between failures) | How long it usually runs before failing | Higher is better |
| MTTR (mean time to repair) | How long a fix usually takes | Lower is better |
RTO is how long until you’re back. RPO is how much you lost. I keep mine at zero by never doing any work. — Nyx
Avoiding single points of failure
- First-hop redundancy (FHRP): two routers share one virtual gateway IP and MAC. VRRP is the open standard; HSRP is Cisco’s. When the active router dies, the standby takes over the address and announces it with gratuitous ARP, which you saw in our ARP lab.
- Link aggregation (LACP, 802.3ad / 802.1AX): several cables act as one, for bandwidth and for survival if one fails. On servers this is called NIC teaming or bonding.
- Active-active vs active-passive: both units carry traffic, or one waits on standby.
- Power: UPS for minutes, generator for hours, dual power supplies fed from different circuits.
Recovery sites
| Site | Ready to run | Cost |
|---|---|---|
| Hot | Within minutes to hours: hardware and current data in place | Highest |
| Warm | Within hours to days: hardware ready, data needs restoring | Medium |
| Cold | Days or more: space and power only | Lowest |
| Cloud site | Depends on the plan: spin up when needed | Pay as you go |
Backups
| Type | Backs up | Restore needs |
|---|---|---|
| Full | Everything | Just the last full |
| Incremental | Changes since the last backup of any type | The last full plus every incremental since |
| Differential | Changes since the last full | The last full plus the latest differential |
The 3-2-1 rule: 3 copies of your data, on 2 different kinds of storage, with 1 copy off-site (or offline, so ransomware can’t reach it). A backup you’ve never restored is a hope, not a backup: test restores on a schedule. Test the whole plan with a tabletop exercise (walk through a scenario on paper) before a real outage tests it for you.
Documentation that pays for itself
| Document | Answers |
|---|---|
| Physical diagram | Where is everything, which cable goes where, which rack and port |
| Logical diagram | Subnets, VLANs, routing, how traffic flows |
| Rack diagram and cable map | What is in each rack unit and patch-panel port |
| IP address management (IPAM) | Which addresses are used, reserved and free |
| Asset inventory and life cycle | What you own, warranty, end-of-life and end-of-support dates |
| Baseline configuration and performance baseline | What “normal” looks like, so you can spot abnormal |
| SLA | What uptime and response the provider promised |
| Wireless survey / heat map | Where the Wi-Fi coverage is and isn’t |
Change management
Every planned change follows the same loop: request (what and why), assess risk and impact, approve (often a change advisory board), schedule a maintenance window, test, implement with a rollback plan, verify, and document. It’s the troubleshooting method’s cousin: plan, act, verify, write it down.
Nova: Start small. Draw your home network on one page: internet connection, router, switch, access points, and every device with a fixed address. That page is a logical diagram, and you just wrote your first network documentation.
Practise it
These topics are in the Network+ practice test, scored by exam domain.
NextNetwork security explained→
Go deeper
- Troubleshooting capstone — the method, hands-on
- Syslog and SNMP lab — monitoring and baselines in practice
Explainer, written by StashGrid with AI help and checked against the CompTIA Network+ N10-009 objectives and the published standards (IEEE 802.11 and 802.3, TIA-568) before publishing. Speeds and distances are the standard ratings; real-world results are lower. StashGrid isn’t affiliated with CompTIA. Spotted an error? Tell us.