Skip to content

README.md

Latest commit

 

History

History
133 lines (94 loc) · 3.27 KB

README.md

File metadata and controls

133 lines (94 loc) · 3.27 KB
Raw

Lab 8 — Static Routing Between Two Subnets

Routers forward packets between subnets using a routing table. In this lab you will build a tiny three-node topology: subnet A ↔ Router ↔ subnet B, and add the static routes that make A and B talk.

Run on https://killercoda.com/playgrounds/scenario/ubuntu

Required software (free, pre-installed):

  • iproute2
  • Linux kernel sysctl (built in)

Files shipped with this lab

File Use
setup.sh Builds the whole topology (Steps 1–5) in one go
verify.sh Objective PASS/FAIL check of Step 6
cleanup.sh Reverses Step 7 and sweeps leftover veth halves
data/routing-table.csv Routing-table planning worksheet used in Step 6
checklist.md Tick-box list of everything you must be able to show

Work through Steps 1–7 by hand the first time — that is where the learning is. If you need to rebuild the topology quickly (e.g. after a playground reset), run ./setup.sh, which executes exactly the commands in Steps 1–5.

Step 1 — Create three namespaces

ip netns add a
ip netns add r1
ip netns add b

r1 will be our router.

Step 2 — Wire A↔R1 and B↔R1 with veth pairs

ip link add a-r type veth peer name r-a
ip link add b-r type veth peer name r-b
ip link set a-r netns a
ip link set r-a netns r1
ip link set b-r netns b
ip link set r-b netns r1

Step 3 — Assign IPs (two different subnets)

ip netns exec a  ip addr add 10.1.1.2/24 dev a-r
ip netns exec r1 ip addr add 10.1.1.1/24 dev r-a
ip netns exec r1 ip addr add 10.2.2.1/24 dev r-b
ip netns exec b  ip addr add 10.2.2.2/24 dev b-r
for n in a r1 b; do ip netns exec $n ip link set lo up; done
ip netns exec a  ip link set a-r up
ip netns exec r1 ip link set r-a up
ip netns exec r1 ip link set r-b up
ip netns exec b  ip link set b-r up

Step 4 — Enable IP forwarding on the router

A router must explicitly forward packets between interfaces.

ip netns exec r1 sysctl -w net.ipv4.ip_forward=1

Step 5 — Add static routes on the end hosts

Each host knows only its own subnet, so we tell it to send everything else through R1.

ip netns exec a ip route add 10.2.2.0/24 via 10.1.1.1
ip netns exec b ip route add 10.1.1.0/24 via 10.2.2.1

Step 6 — Verify

ip netns exec a ping -c2 10.2.2.2
ip netns exec a ip route

Compare the output of ip route with the planning worksheet data/routing-table.csv, which lists every destination, mask, gateway, interface and metric in this topology plus a few extra subnets to plan for:

cat data/routing-table.csv

Then run the automated check, which asserts the addressing, ip_forward, both static routes and the end-to-end ping:

./verify.sh

Step 7 — Clean up

for n in a r1 b; do ip netns del $n; done

Or run the provided script, which does the same and also removes any veth half left behind:

./cleanup.sh

Finally, work through checklist.md and tick off each outcome.

What you learned

  • Routing requires net.ipv4.ip_forward=1 on the router.
  • End hosts use static routes (or a default gateway) to reach networks beyond their LAN.
  • The format via <next-hop> tells the kernel which neighbour to send packets to.