// Practice lab
CCNA lab 12: Design a VLSM addressing plan
Scenario
A company has been given 10.40.0.0/22 for two sites. Plan every subnet, largest first, so that each segment has room for about 25% growth.
| Site | Segment | Hosts today |
|---|---|---|
| HQ | Staff | 300 |
| HQ | Voice | 150 |
| HQ | Wi-Fi guests | 100 |
| HQ | Servers | 40 |
| Branch | Staff | 60 |
| Branch | Voice | 30 |
| Both | Network management | 20 |
| WAN | HQ–Branch link | 2 |
| WAN | HQ–Internet firewall link | 2 |
Tasks
- Add 25% to each host count and choose the smallest prefix that fits.
- Allocate the subnets from 10.40.0.0, largest first.
- For each subnet write the network, usable range, broadcast and the gateway (first usable address).
- How much of the /22 is left over?
- Which subnets could be summarised for the branch?
Show the solution
Step 1 — requirements with growth:
| Segment | Hosts × 1.25 | Prefix | Size |
|---|---|---|---|
| HQ Staff | 375 | /23 (510) | 512 |
| HQ Voice | 188 | /24 (254) | 256 |
| HQ Wi-Fi | 125 | /25 (126) | 128 |
| Branch Staff | 75 | /25 (126) | 128 |
| HQ Servers | 50 | /26 (62) | 64 |
| Branch Voice | 38 | /26 (62) | 64 |
| Management | 25 | /27 (30) | 32 |
| WAN HQ–Branch | 2 | /30 | 4 |
| WAN HQ–Firewall | 2 | /30 | 4 |
Total: 512 + 256 + 128 + 128 + 64 + 64 + 32 + 4 + 4 = 1192 of 1024 addresses — it does not fit. This is the real lesson: check the total before allocating. Options: reduce growth for some segments, or request a larger block. Using the counts without growth for HQ Staff (300 → /23 is still needed) and HQ Voice (150 → /24) doesn't help; the /23 alone is half the block.
A workable plan with a /21 (10.40.0.0/21) instead:
| Segment | Network | Usable range | Broadcast | Gateway |
|---|---|---|---|---|
| HQ Staff | 10.40.0.0/23 | 10.40.0.1 – 10.40.1.254 | 10.40.1.255 | 10.40.0.1 |
| HQ Voice | 10.40.2.0/24 | 10.40.2.1 – .254 | 10.40.2.255 | 10.40.2.1 |
| HQ Wi-Fi | 10.40.3.0/25 | 10.40.3.1 – .126 | 10.40.3.127 | 10.40.3.1 |
| Branch Staff | 10.40.3.128/25 | 10.40.3.129 – .254 | 10.40.3.255 | 10.40.3.129 |
| HQ Servers | 10.40.4.0/26 | 10.40.4.1 – .62 | 10.40.4.63 | 10.40.4.1 |
| Branch Voice | 10.40.4.64/26 | 10.40.4.65 – .126 | 10.40.4.127 | 10.40.4.65 |
| Management | 10.40.4.128/27 | 10.40.4.129 – .158 | 10.40.4.159 | 10.40.4.129 |
| WAN HQ–Branch | 10.40.4.160/30 | .161 – .162 | .163 | — |
| WAN HQ–Firewall | 10.40.4.164/30 | .165 – .166 | .167 | — |
Left over: 10.40.4.168 – 10.40.7.255 (about 3.3 /24s) for future sites.
Summarisation: as allocated "largest first", the branch subnets (10.40.3.128/25 and 10.40.4.64/26) are not contiguous, so they cannot be summarised into one route. If summarisation matters, allocate per site instead — for example give the branch its own /24 (10.40.5.0/24) and subnet that: staff 10.40.5.0/25, voice 10.40.5.128/26. Then HQ only needs one route, 10.40.5.0/24, for the whole branch. Real designs balance "largest first" against "one block per site".