Welcome to the generic SR Linux lab, where you will learn and experience Nokia's SR Linux network operating system.
In this lab you can:
- Deploy the pre-configured SR Linux fabric with workloads using containerlab.
- Explore the underlay routing with eBGP and a typical BGP EVPN overlay setup leveraging Route Reflectors.
- Examine the IP-VRF and MAC-VRF network instances, VXLAN tunnels, and IRB interfaces.
- Verify VXLAN datapath by pinging between the hosts.
- And discover everything you want to know about SR Linux!
Difficulty: Beginner
The lab represents a 2-Tier Clos data center fabric topology with four hosts.
The Leaf and Spine nodes are running containerized Nokia SR Linux (7220 IXR D3L) NOS and the hosts are based on Alpine Linux.
Topology diagram:
Every node in the data center fabric is configured with eBGP as an underlay routing protocol. iBGP is used to setup EVPN-based overlay layer 2 and layer 3 services.
Hosts are connected to the respective overlay services as per the diagram below:
Deploy the topology with containerlab.
cd $HOME/DCFWorkshop/srl-generic-lab
sudo clab deploy -c -t srl-generic.clab.ymlIn this lab, you may choose to build out the fabric from scratch or explore a fabric that will fully configured.
If you choose to build from scratch, you will find step-by-step SRLinux configuration examples at Step-by-Step Guide
The default setting is to build from scratch. If you choose to explore a configured fabric, please uncomment the line for the startup-config in the containerlab topology file srl-generic.clab.yml for all 6 nodes.
#startup-config: configs/fabric/leaf1.cfg
Once the lab is running, you can access the network elements from inside the VM through ssh, just passing the node name:
ssh admin@clab-srl-generic-leaf1
# password: NokiaSrl1!
To access host h1:
# password: srllabs@123
sudo docker exec -it clab-srl-generic-h1 bash
The role of the underlay routing is to distribute loopback prefixes across the leaf switches of our fabric. In this lab, we are using eBGP as the underlay routing protocol.
You can connect to nodes and check the interface configs, eBGP sessions state, AS numbers, and underlay network reachability between nodes by pinging the system IP addresses.
A list of helpful show commands:
show interface briefshow interface ethernet-1/3*show network-instance summaryshow network-instance default route-table allshow network-instance default protocols bgp neighborshow network-instance default protocols bgp neighbor 100.64.1.1 advertised-routes ipv4show network-instance default protocols bgp neighbor 100.64.1.1 received-routes ipv4
info commands allow the retrieval of the configuration and state information from the respective datastores and provide full access to the system internals.
Getting information from the configuration datastore:
info interface ethernet-1/*info network-instance default
And from the state datastore:
info from state interface ethernet-1/31 subinterface 1 statisticsinfo from state network-instance default protocols bgp neighbor * received-messages
The output can be retrieved in different formats:
As JSON:
info from state network-instance default protocols bgp neighbor * | filter fields admin-state session-state | as json
As Table:
info from state network-instance default protocols bgp neighbor * | filter fields admin-state session-state | as table
These commands can be associated to CLI Aliases
-environment alias mybgp "info from state network-instance default protocols bgp neighbor * | filter fields admin-state session-state | as table"
-mybgp
These CLI Aliases can receive inputs so that the neighbor can be dynamically selected (dynamic aliases)
-environment alias mybgp "info from state network-instance default protocols bgp neighbor {} | filter fields admin-state session-state | as table"
-mybgp mybgp 10.0.0.5
EVPN uses MP-BGP as a control plane protocol between the tunnel endpoints. Typically, route-reflectors (RRs) are configured for the scalability. In this lab, spine routers are configured as RRs. All the leaf routers are peering with the spine routers for iBGP EVPN routes.
Some useful commands:
show network-instance default protocols bgp neighbor 10.0.0.5 advertised-routes evpnshow network-instance default protocols bgp neighbor 10.0.0.5 received-routes evpninfo network-instance default protocols bgp group ibgp-evpninfo from state network-instance default protocols bgp neighbor 10.0.0.5 peer-type
MAC-VRFs are layer 2 EVPN instances usually mapped to a single broadcast domain and forward based on the mac-table. IP-VRFs are layer 3 instances with IP routing tables. MAC-VRFs can bind to IP-VRFs using IRB interfaces.
Connect to the SR Linux nodes to check the EVPN service building blocks; IP-VRF and MAC-VRF configurations, VXLAN tunnel and IRB interfaces.
Have a look at the overlay topology and how different VRF types are mapped:
Some useful commands:
info network-instance ip-vrf-1info from state network-instance mac-vrf-1 protocolsshow network-instance mac-vrf-1 bridge-table mac-table allshow network-instance ip-vrf-1 route-table allinfo tunnel-interface vxlan1 vxlan-interface *show tunnel-interface vxlan1 vxlan-interface * detailinfo interface irb1info interface ethernet-1/1*show network-instance default protocols bgp routes evpn route-type summary
With overlay services deployed, you can successfully test network reachability between the hosts.
Connect to the hosts and ping from:
- h1 to h4 (ping goes over mac-vrf-1)
sudo docker exec -it clab-srl-generic-h1 ping 100.101.1.14
- h1 to eth1 of h3 (ping goes over ip-vrf-1)
sudo docker exec -it clab-srl-generic-h1 ping 100.101.3.13
- h2 to eth2 of h3 (ping goes over ip-vrf-2)
sudo docker exec -it clab-srl-generic-h2 ping 100.102.4.13
- h3 to 100.100.100.100 (ping an address on ip-vrf-1 and ip-vrf-2)
sudo docker exec -it clab-srl-generic-h3 ping 100.100.100.100
Check the mac/ip table entries and EVPN route advertisements on the related SR Linux nodes with:
show network-instance mac-vrf-1 bridge-table mac-table allshow network-instance ip-vrf-1 route-table allshow network-instance default protocols bgp neighbor 10.0.0.5 received-routes evpnshow network-instance default protocols bgp routes evpn route-type summaryshow network-instance default protocols bgp routes evpn route-type 2 detailshow network-instance default tunnel-table ipv4show tunnel-interface vxlan1 vxlan-interface * detailinfo from state tunnel vxlan-tunnel vtep * statistics