How many usable IPs does a /30 actually give you?
You’re staring at a subnet mask—255.252—and wondering why the calculator keeps spitting out “4 IPs, 2 usable.255.” It feels like one of those networking riddles that only admins in dark server rooms get to solve. 255.The short answer is “two,” but the story behind that number is worth a few minutes of your time And that's really what it comes down to..
What Is a /30 Subnet
A /30 is a block of IPv4 addresses where the first 30 bits are fixed as the network portion. In plain English, you’ve taken the 32‑bit address space and borrowed two bits for hosts. Those two bits can be either 00, 01, 10, or 11, which gives you four address slots total.
Network address
The all‑zeros combination (…00) is the network identifier. It tells routers “this is the subnet itself,” not a device you can ping.
Broadcast address
The all‑ones combination (…11) is the broadcast address. Anything sent to it gets delivered to every host in that little subnet Worth keeping that in mind. Nothing fancy..
Host addresses
The remaining two combos (…01 and …10) are the only ones you can assign to actual devices—your servers, routers, or whatever you’re wiring up And that's really what it comes down to..
That’s why you end up with two usable IPs out of the four that the /30 block technically contains It's one of those things that adds up. That's the whole idea..
Why It Matters / Why People Care
If you’ve ever set up a point‑to‑point link between two routers, you’ve probably been handed a /30. The idea is simple: each end of the link needs its own IP, and you don’t want any extra addresses floating around that could cause confusion Took long enough..
Real‑world impact
- Avoiding IP conflicts – Using a /30 guarantees no other device on the same LAN will accidentally grab the broadcast or network address.
- Conserving address space – In an era where IPv4 is getting scarce, squeezing the smallest possible subnet out of a block can save you a lot of headaches later.
- Simplifying routing tables – A /30 creates a clean, predictable route entry on each router, which makes troubleshooting easier.
When you misunderstand the count and try to assign three devices to a /30, you’ll hit the dreaded “duplicate IP” error or lose connectivity because one of those devices is actually the broadcast address. That’s why the “2 usable IPs” rule isn’t just trivia; it’s a practical guardrail.
How It Works (or How to Do It)
Let’s walk through the math and the steps you’d actually follow when you need a /30 for a link.
1. Identify the network block
Assume you’ve been given the block 192.On the flip side, 168. 10.255.But 0/30. Which means the mask 255. 255.252 tells you the last octet only has two bits for hosts.
2. List the four addresses
| Binary (last octet) | Decimal | Meaning |
|---|---|---|
| 00 | 0 | Network address |
| 01 | 1 | First usable host |
| 10 | 2 | Second usable host |
| 11 | 3 | Broadcast address |
So the full list is:
- 192.168.10.0 – network
- 192.168.10.1 – usable #1
- 192.168.10.2 – usable #2
- 192.168.10.3 – broadcast
3. Assign the host IPs
Typically you’ll give the lower usable IP to the “local” router and the higher one to the “remote” router. Example:
- Router A: 192.168.10.1/30
- Router B: 192.168.10.2/30
Both routers now know the exact subnet they belong to, and any traffic addressed to 192.Think about it: 168. On top of that, 10. 3 will be treated as a broadcast.
4. Verify with a quick ping test
From Router A:
ping 192.168.10.2 # should get replies
ping 192.168.10.3 # should get no replies (broadcast)
If the first ping works and the second times out, you’ve set it up right.
5. Add the static route (optional)
On each router, you might need a static route pointing to the other side:
ip route 10.0.0.0/24 192.168.10.2 # on Router A
ip route 10.0.1.0/24 192.168.10.1 # on Router B
That tells the routers how to reach the rest of your network through the point‑to‑point link Most people skip this — try not to. Which is the point..
Common Mistakes / What Most People Get Wrong
- Counting the broadcast as usable – Newbies often think “four IPs, so four devices.” The broadcast can’t be assigned to a host; it’s a special address.
- Using the network address for a device – Plugging the network ID into a server will make it invisible to the rest of the world.
- Assuming /30 works for LANs – A /30 is great for a link between two points, but trying to run a whole office LAN on it will leave you with only two computers!
- Forgetting the mask on the device – If you set the IP but leave the subnet mask at the default /24, the device will think the whole 255.255.255.0 network is reachable, breaking the point‑to‑point logic.
- Over‑allocating IPs from a larger block – Some admins grab a whole /24 just to get two IPs, then waste the rest. That’s a classic IPv4‑conservation faux pas.
Practical Tips / What Actually Works
- Use a calculator – Even a quick mental check can go wrong. Online CIDR calculators (or a simple spreadsheet) will instantly show you the network, usable, and broadcast addresses.
- Label your diagrams – When you draw the topology, write the network, usable, and broadcast IPs next to each box. It saves you from mis‑wiring later.
- Reserve the .0 and .255 wisely – In many older networks, .0 was treated as “network” and .255 as “broadcast” across the whole classful range. Modern CIDR ignores that, but some legacy gear still does.
- Document the purpose – Note in your change log that the /30 is a point‑to‑point link. Future engineers will thank you when they see a tiny subnet and wonder why it exists.
- Consider IPv6 for new links – If you’re building a fresh environment, you can skip the whole /30 dance and use a /64 IPv6 subnet. It’s abundant and removes the “usable vs. broadcast” confusion entirely.
FAQ
Q: Can I use a /30 for a LAN with three devices?
A: No. A /30 only gives you two host addresses. For three devices you need at least a /29 (8 total, 6 usable).
Q: Why does the broadcast address matter if I’m only using two hosts?
A: Because the broadcast address is reserved by the protocol. If you accidentally assign it, the device will never be reachable and may cause ARP storms Not complicated — just consistent..
Q: Is there any scenario where I’d want more than two usable IPs in a point‑to‑point link?
A: Rarely. Some vendors allow a “/31” for point‑to‑point, which gives you exactly two addresses and treats both as host addresses—no network or broadcast. It’s a newer RFC‑compatible trick.
Q: How do I calculate the next /30 block after 192.168.10.0/30?
A: Add four to the last octet. The next block is 192.168.10.4/30, then 192.168.10.8/30, and so on And that's really what it comes down to..
Q: Can I subnet a /30 further?
A: No. You’ve already borrowed all but two bits for hosts. Splitting it would give you subnets with zero usable addresses, which is pointless The details matter here..
That’s the whole picture: a /30 hands you four addresses, but only two can actually sit on a device. Even so, knowing which two are yours, and why the other two exist, saves you from a lot of “why won’t this ping? ” moments. Next time you see a /30, you’ll be able to assign it confidently, document it clearly, and keep your network tidy—without wasting precious IPv4 real estate. Happy subnetting!
Real-World Deployment Scenarios
When you're actually in the field, /30 subnets show up in predictable places. Understanding these patterns helps you recognize when to use them without overthinking.
WAN and MPLS Links – Most carrier circuits use /30 addressing. The provider owns one end, your edge router owns the other. You rarely need more than two IPs here, which is why /30 became the de facto standard.
Stacked Switches – Some switch stacking protocols use a dedicated /30 between members for control plane traffic. It's isolated from the production VLANs and sized tiny for a reason Worth keeping that in mind..
Loopback Interfaces – Loopbacks are /32 by default (single host), but if you're routing OSPF or BGP between two routers via a transit link, that link is almost certainly a /30.
Management Ports – Out-of-band management often gets its own /30 when you have exactly two devices: a console server and a managed switch or firewall. It's cleaner than cramming management into a larger VLAN.
Automation and Tooling
If you're provisioning links at scale, manual IPAM gets painful. Consider these approaches:
- Terraform with cloud providers – Many providers allocate /30s automatically for VPC peerings or transit gateway attachments. You rarely even see the addresses.
- Ansible templates – Use Jinja2 templates to generate /30 assignments from a list of site codes. Reduces typos and keeps documentation in sync with reality.
- NetBox or IPAM software – These tools track which /30s are free, which are used, and what they're for. At scale, you cannot manage point-to-point links in a spreadsheet.
The Bottom Line
/30 subnets are small, efficient, and purpose-built. Now, they exist because two devices need to talk, and the protocol demands a network address and a broadcast address. That's four IPs, two usable, and zero room for negotiation.
When you understand why the math works, you stop second-guessing yourself. You label the diagram, reserve the block, document the purpose, and move on. The network stays clean, the troubleshooting stops before it starts, and you save the /30s for what they do best—connecting exactly two things Not complicated — just consistent..
That, in the end, is all a /30 ever was meant to be Not complicated — just consistent..
