NetPractice – IP Addressing & Routing Fundamentals 🌐🧠

✅ Status: Completed (10/10 levels exported)
🏫 School: 42 – NetPractice
🏅 Score: 100/100

A hands-on networking sandbox: subnetting, gateways, and routing — with instant feedback.

---

📚 Table of Contents

• 📝 Description

• 🎯 What You Learn

• 🚀 Quick Start

• 🧭 How NetPractice Thinks (Mental Model)

  • Layer 2 vs Layer 3
  • Interfaces = Subnet borders
  • “Same subnet?” rule

• 🧩 Subnetting Cheat Sheet

  • Subnetting from Bits (Calculator‑friendly)
  • CIDR → Mask (the bit view)
  • Mask → CIDR (fast)
  • “Step / Block size”
  • Why network boundary IPs appear in routes
  • IP AND MASK = NETWORK
  • Network / Broadcast: quick boundary rules
  • Fast host validity checklist
  • CIDR → Mask (Fast Table)
  • Block Size Method (Fastest Way)
  • Network / Broadcast / Host Range
  • Common CIDR Patterns in NetPractice

• 📏 How Many Addresses a Mask Gives You

  • CIDR /n → total addresses and usable hosts
  • Dotted mask → how many addresses?
  • /30 explained

• 🧷 Reserved IPv4 Ranges

• 🛣️ Routing Cheat Sheet

  • Default Gateway
  • Static Routes
  • Longest Prefix Match
  • Route Summarization

• 🧰 A Practical Workflow to Solve Any Level

• 🧯 Debugging & Common Mistakes

  • Interpreting the simulator logs
  • Why “loop detected” happens

• 📦 Submission Files

• 📂 Repository Layout

• 🔗 Resources

• 🧾 42 Notes (Defense Tips)

---

📝 Description

NetPractice is an interactive networking project where you fix 10 broken network topologies by configuring:

• IPv4 addresses
• Subnet masks (CIDR)
• Default gateways
• Router routes (static routing)

Each level provides goals and a live packet simulation log, so you can validate your changes instantly.

---

🎯 What You Learn

By the end, you’ll be comfortable with:

• ✅ Understanding subnets and CIDR (/24, /26, /30, …)
• ✅ Calculating network address, broadcast, and host ranges
• ✅ Recognizing what belongs to Layer 2 vs Layer 3
• ✅ Setting correct default gateways
• ✅ Adding static routes to reach non-direct networks
• ✅ Reading logs and reasoning about where packets break

---

🚀 Quick Start

1. Extract the NetPractice archive.
2. Open index.html in a browser (Chrome/Chromium recommended).
3. Solve levels 1 → 10.
4. For each level, click Get my config and export your file.
5. Put all 10 exported files into the root of this repository.

---

🧭 How NetPractice Thinks (Mental Model)

Layer 2 vs Layer 3 (super important)

• Switch (L2): forwards frames inside a LAN. No routing logic.
• Router (L3): forwards packets between networks. Chooses next hop by routing table.

Interfaces = Subnet borders

A router interface belongs to exactly one subnet. So whenever you see a link from R1 to a host/switch segment:

• the router interface IP must be inside that segment’s subnet
• hosts in that segment must also be inside that subnet

“Same subnet?” rule

Two devices can communicate directly only if:

• their IPs are in the same subnet (based on the mask)
• and the link is physically connected

If not, you need a router (gateway) in between.

---

🧩 Subnetting Cheat Sheet

🧠 Subnetting from Bits (Calculator‑friendly)

When you only have a basic calculator, think in bits → bytes → decimals.

1) CIDR → Mask (the bit view)

A CIDR /n means: n network bits, then the rest are host bits.

• Split n into full octets + remainder:

  • full = n / 8
  • rem = n % 8

• The first full octets are 255.

• The next octet is built from the first rem bits set to 1.

Remainder → octet value table:

rem bits	binary	value
0	00000000	0
1	10000000	128
2	11000000	192
3	11100000	224
4	11110000	240
5	11111000	248
6	11111100	252
7	11111110	254
8	11111111	255

Example: /26

• 26 = 3×8 + 2 → 255.255.255.(192) → 255.255.255.192

2) Mask → CIDR (fast)

• Count 255 octets: each 255 = 8 bits
• Then add bits from the table above for the first non‑255 octet

Example: 255.255.255.224

• three 255 → 24 bits
• 224 corresponds to rem=3 → total /27

3) “Step / Block size” (what we called «шаг»)

If subnetting happens inside an octet, compute:

• block = 256 − mask_octet

Then subnet starts in that octet are:

• 0, block, 2×block, 3×block, ...

This is the fastest method you and I used across levels.

Example: /26 → last octet mask is 192 → block = 64 → starts: 0, 64, 128, 192.

4) Why “network boundary” IPs appear in routes (like .128)

A route destination is a network, not a host. So writing 129.128.239.128/25 is valid even if no device has .128. It simply means “the subnet that starts at .128”.

5) The real math: IP AND MASK = NETWORK

This is the bit-level truth behind everything.

• Convert IP and mask to binary (mentally you usually only need the changing octet).

• Apply AND:

  • 1 AND 1 = 1
  • 1 AND 0 = 0
  • 0 AND 1 = 0
  • 0 AND 0 = 0

Example: 192.168.10.130/26

• changing octet: 130
• /26 → mask last octet = 192

Binary:

• 130 = 10000010
• 192 = 11000000

AND:

• 10000010
• 11000000 = 10000000 = 128

So the network is 192.168.10.128.

Why we still use “block size”: it’s the same idea, just faster than doing AND every time.

6) Network / Broadcast: quick boundary rules

Once you know the block size and subnet start:

• Network = subnet start
• Broadcast = next subnet start − 1
• Valid hosts = between them (excluding both ends)

Example: ...128/26

• next subnet start is ...192
• broadcast is ...191
• hosts: ...129 → ...190

7) Fast host validity checklist

Before you even look at routing, verify:

• ✅ IP is inside the subnet range
• ✅ IP is not the network address
• ✅ IP is not the broadcast address
• ✅ Gateway (if needed) is in the same subnet

---

CIDR → Mask (Fast Table)

CIDR	Mask	Hosts (usable)	Typical use
/24	255.255.255.0	254	classic LAN
/25	255.255.255.128	126	split /24 in 2
/26	255.255.255.192	62	split /24 in 4
/27	255.255.255.224	30	small LAN
/28	255.255.255.240	14	very small LAN
/29	255.255.255.248	6	tiny LAN
/30	255.255.255.252	2	point-to-point link

Quick host count formula: usable hosts = 2^(32−CIDR) − 2

---

Block Size Method (Fastest Way)

When the mask breaks inside an octet, compute block size:

• block = 256 − mask_octet

Examples:

• /26 → mask 255.255.255.192 → block 256 − 192 = 64
• /27 → mask 255.255.255.224 → block 256 − 224 = 32
• /30 → mask 255.255.255.252 → block 256 − 252 = 4

This block is the step between subnet starts in the changing octet.

---

Network / Broadcast / Host Range

If your changing octet is, say, the last one:

• subnet starts at: 0, block, 2×block, 3×block, ...
• Network = nearest subnet start ≤ IP
• Broadcast = next subnet start − 1
• Host range = (network + 1) ... (broadcast − 1)

Example: 192.168.10.130/26

• /26 → block size 64 → subnet starts: 0, 64, 128, 192
• 130 is inside 128–191
• Network: 192.168.10.128
• Broadcast: 192.168.10.191
• Hosts: 192.168.10.129 – 192.168.10.190

---

Common CIDR Patterns in NetPractice

/30 (router-to-router / point-to-point)

• Block size 4 in last octet

• Each subnet has exactly:

  • network
  • 2 usable IPs
  • broadcast

Example subnet: 10.0.0.8/30

• network: 10.0.0.8
• usable: 10.0.0.9 and 10.0.0.10
• broadcast: 10.0.0.11

Typical /30 mistakes

• ❌ Using network or broadcast as an interface IP
• ❌ Trying to place 3+ devices into a /30 (it only supports 2 usable IPs)
• ❌ Setting next-hop to an IP that is not one of the two usable addresses

/26 (subnetting practice)

• One /24 becomes 4 subnets:

  • .0/26, .64/26, .128/26, .192/26

This is exactly why you often see addresses like .128 or .192: they are subnet boundaries, not necessarily host IPs.

---

📏 How Many Addresses a Mask Gives You

This is one of the fastest ways to sanity-check your configs.

1) CIDR /n → total addresses and usable hosts

• Host bits: h = 32 − n

• Total addresses in the subnet: 2^h

• Usable hosts (classic subnet): 2^h − 2

  • one is the network address (all host bits = 0)
  • one is the broadcast address (all host bits = 1)

Quick examples:

CIDR	Host bits h	Total addresses	Usable hosts
/24	8	256	254
/26	6	64	62
/30	2	4	2

Note: Some real-world networks use /31 for point-to-point (RFC 3021) and /32 for a single host route. In NetPractice you’ll mostly see the classic “usable = total − 2” rule.

2) Dotted mask (e.g., 255.255.255.0) → how many addresses?

• Each 255 means that octet has 0 host bits (it’s fully network bits).
• The first non-255 octet tells you the block size and the number of addresses per subnet.

Examples:

• 255.255.255.0 is /24

  • host bits = 8 → total 256 → usable 254

• 255.255.255.252 is /30

  • host bits = 2 → total 4 → usable 2

3) /30 explained (why only 2 usable IPs)

A /30 has 4 addresses:

• network (…0)
• host (…1)
• host (…2)
• broadcast (…3)

That’s why it’s perfect for router-to-router links: one IP per router interface.

---

🧷 Reserved IPv4 Ranges (Private, Local, Special)

These ranges are not “normal public internet” addresses.

Private (RFC1918) — used inside LANs

• 10.0.0.0/8
• 172.16.0.0/12 (172.16.0.0 → 172.31.255.255)
• 192.168.0.0/16

NetPractice may warn you with messages like “private subnets not routed over internet” if you try to send these through an “Internet cloud”.

Link-local (APIPA) — local segment only

• 169.254.0.0/16

Used when a host can’t get DHCP; traffic stays on the local link (not routed).

Loopback — the host itself

• 127.0.0.0/8 (commonly 127.0.0.1)

Carrier‑Grade NAT (CGNAT) — ISP internal space

• 100.64.0.0/10

Not public end-user space; used by ISPs for large-scale NAT.

Multicast

• 224.0.0.0/4 (224.0.0.0 → 239.255.255.255)

Broadcast / “this network” special cases

• 255.255.255.255 — limited broadcast
• 0.0.0.0 — “unspecified / default route” (you’ll see 0.0.0.0/0 as the default route)

In NetPractice, the big takeaway is simple:

• Use private ranges for internal LAN segments unless the level forces public-looking IPs.
• Don’t treat special ranges (loopback, link-local, multicast) as normal host LAN addressing.

---

🛣️ Routing Cheat Sheet

Default Gateway

A host sends traffic like this:

1. If destination is in same subnet → send directly.
2. Otherwise → send to default gateway.

✅ The default gateway must be:

• in the same subnet as the host
• the IP of the router interface connected to that LAN

---

Static Routes

A router knows only:

• networks directly connected to its interfaces
• routes you explicitly add (static routes)
• (no dynamic routing in NetPractice)

A static route conceptually says:

“To reach DEST/MASK, forward to NEXT_HOP.”

✅ NEXT_HOP must be:

• reachable via one of the router’s directly-connected networks

---

Longest Prefix Match (Route Choice)

Routers choose the most specific matching route:

• /24 is more specific than /16
• /26 is more specific than /24

So if both exist:

• 10.0.0.0/8 and 10.42.0.0/16

Traffic to 10.42.5.10 will use 10.42.0.0/16.

---

Route Summarization (Supernetting)

Sometimes one route can cover multiple subnets. Typical NetPractice trick:

• combine multiple contiguous subnets into a bigger prefix

Example:

• 192.168.10.128/26 and 192.168.10.192/26
• can be summarized as 192.168.10.128/25

That’s why you might write a network boundary address (like .128) even if no device has exactly that IP — it represents the subnet itself.

---

🧰 A Practical Workflow to Solve Any Level

Use this every time:

1. Read goals first (what exact connectivity is required).

2. Pick one goal pair (A → B) and trace the path.

3. For every link segment:

   • ensure endpoints are in the same subnet

4. For every host:

   • ensure IP/mask valid
   • if talking outside subnet → default gateway set and reachable

5. For every router:

   • interfaces must be in correct subnets
   • add routes for remote networks

6. Re-run Check again and read logs.

Pro tip: fix one direction first (A→B), then check reverse (B→A).

---

🧯 Debugging & Common Mistakes

🧾 Interpreting the simulator logs (what they usually mean)

You can often solve a level just by reading the log like a story.

• "destination does not match any interface"

  • The device is a host and the destination isn’t in its subnet → it will try the routing table.
  • If it then uses 0.0.0.0/0, your default route/gateway is in play.

• "route match 0.0.0.0/0"

  • No more specific route existed → device used default route.
  • If it fails afterward, your gateway is wrong / unreachable.

• "send to gateway X through interface Y"

  • Device decided to forward to next-hop X using interface Y.
  • If X is not in the same subnet as Y, it can’t ARP / can’t deliver.

• "packet not for me" (on router)

  • Router received it, checked destination, and will forward based on routes.

• "loop detected"

  • The packet bounced back to a device it already visited.

🔁 Deep dive: Why “loop detected” happens (and how to fix it fast)

✅ Important NetPractice note: a switch can trigger “loop detected” in logs

In NetPractice logs you may see sequences like:

• on switch S: pass to all connections
• on switch S test link to A
• on A: loop detected

This can happen because the simulator floods / tests links on a switch-like device, and a packet may appear to come back to a node that already saw it.

✅ If you still see "destination IP reached" and all goals are valid, this “loop detected” is often a simulator artifact rather than a real network loop.

Still, treat it as a quick signal to double-check:

• host gateways are in the same subnet
• return routes exist on routers
• default routes don’t point to each other

Think of a packet like this:

If a device doesn’t know the exact path, it sends to its default route. If two devices “default” to each other (or a router sends back where it came from), the packet will ping‑pong → loop.

The 3 most common loop patterns in NetPractice

1. Wrong host gateway

• Host A wants to reach remote network.
• A uses gateway G, but G is not the router interface for that LAN (or not even reachable).
• The frame gets flooded / bounced, and the simulator detects a loop.

✅ Fix:

• Gateway must be the router interface IP on the same LAN as the host.

2. Missing route on a router (return path problem)

• Forward direction works (A → B), because router knows where B is.
• Reverse direction fails (B → A), because some router in the path has no route back.
• Packet goes to default route and comes back → loop.

✅ Fix:

• Always validate both directions.
• Add the missing static route(s) so every router knows how to reach the source subnet.

3. Two routers defaulting to each other

• Router R1 has default route → R2
• Router R2 has default route → R1
• Any unknown destination causes infinite ping‑pong.

✅ Fix:

• Avoid mutual default routes.
• Add specific routes (more specific prefix) or make only one side default to the other, and the other side has explicit routes.

Quick “Loop detected” checklist (30 seconds)

• ✅ Is the gateway on each host in the same subnet?
• ✅ Does each router have a route to the destination subnet?
• ✅ Does each router have a route back to the source subnet? (return path!)
• ✅ Are any two routers pointing 0.0.0.0/0 at each other?
• ✅ Is any next-hop IP reachable via a directly connected subnet?

Pro tip: if logs show a device twice in the path (A → … → A), the loop is usually between the last two hops before it repeats.

• "private subnets not routed over internet"

  • You are trying to route RFC1918 private ranges (e.g. 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16) “through the internet cloud”.
  • Fix: keep private traffic inside your controlled routers or use the public addressing in the exercise.

Common mistakes

• ❌ Gateway is set but not in the same subnet as the host
• ❌ Router route points to a next hop that is not reachable
• ❌ Two devices look “close” but are in different subnets due to mask
• ❌ Using a broadcast or network address as a host IP
• ❌ Forgetting that /30 has only 2 usable IPs

How to read logs

Logs usually tell you exactly where the packet died:

• accepted / not for me
• destination does not match any interface
• route match 0.0.0.0/0 (default route)
• loop detected

When you see loop detected:

• you probably have a wrong gateway or missing route causing packets to bounce.

---

📦 Submission Files

This repository contains:

• 10 exported config files (one per level) in the repository root
• README.md

Double-check exported filenames before pushing.

---

📂 Repository Layout

.
├── README.md
├── level1.*
├── level2.*
├── level3.*
├── level4.*
├── level5.*
├── level6.*
├── level7.*
├── level8.*
├── level9.*
└── level10.*

(Extensions depend on your export format.)

---

🔗 Resources

• RFC 791 — IPv4
• RFC 4632 — CIDR
• “Subnetting” basics (any networking textbook)
• OSI vs TCP/IP model overviews

Optional tools for practice (outside defense):

• any subnet calculator website (just for training, not for evaluation)

---

🧾 42 Notes (Defense Tips)

• Expect to solve random levels during defense.
• External tools are generally not allowed; a simple calculator like bc may be allowed.
• Practice mental subnetting: /26, /27, /30 should become instant.

---

If you want, I can also add a mini “level cookbook” section with typical patterns you’ll see (e.g., “2 LANs + 1 router”, “3 routers chain”, “summarize /26 blocks into /25”), based on your exported solutions — you can paste screenshots or describe a couple of your hardest levels.