MPLS in the SDN Era --> DevNet SPAUTO
All of the study notes have now been moved to use auto-generated documentation to build a static site with Github Pages. Alongside all the code documentation, you will find a folder structure for notes in markdown broken down by test categories.
A lab topology based on MPLS in the SDN era book used for 300-535 SPAUTO studies
Cisco Certified DevNet Specialist - Service Provider Automation and Programmability
Why MPLS in the SDN Era?
Simple. This is an incredible book for any and all Network Engineers interested in learning technologies used in Service Provider environments. The book is heavily focused on MPLS, SDN, Segment-Routing, BGP, L2VPN/L3VPNs, over-all traffic engineering and much more. These are common terms within SP networks and the book does an incredible job in guiding the readers while exploring the device configurations in this topology. However, many other topics are not covered by the book and this REPO will hopefully fill in the gaps. This book lays the foundation for our lab topology, but it has increased and grown over time. Other material that has been helpful in studying is the Network Programmability with YANG book, which does cover a lot of the topics in the blueprint!
This book goes through a number of device configurations in a multi-vendor lab topology (Cisco IOSXR && Juniper). However, For this lab, all devices have been replaced with Cisco-IOSXR/XE as we are working on a Cisco Certification.
To fully automate Service Provider Networks, you must understand Service Provider Networks
Network Programmability with YANG
What does this repository include & cover
My goal is to gather all of my self-training in this repository and cover all the major Cisco Certified DevNet Specialist (300-535 SPAUTO) topics.
Extras
There will be some extra material in this repository that can be an aid for other areas including but not limited to:
- Batfish
- Nornir
How can I use this repo to study?
I recommend you read the Documentation for my study notes and if you want to play around creating scripts with Nornir, Ansible using SSH/NETCONF, etc. Go ahead and start the topology using ContainerLab and have fun automating the network!
30% of the SPAUTO exam is around Automation and Orchestration platforms, such as NSO. I can't stress enough the importance of taking the time to setup an NSO instance to explore and take advantage of the many examples in this repository. Containerizing NSO for CI/CD is also part of the blueprint, which we will cover.
Why rely on this complicated lab topology?
I learn better when I am doing, not just reading and specially when breaking things. Having this complicated topology, building all the services in an automated way will be key to our success in passing the certification. There will be tons of material in the end that will cover all the topics in the exam.
Study Material & Resources
- MPLS in the SDN Era
- Network Programmability with YANG
- Nicholas Russo DEVNET Material (Pluralsight)
Container Lab
After exploring ContainerLab, I have begun to move away from EVE-NG and completely dockerize this entire topology. The default branch on this project is now containerlab
. Initially, I thought I would share an exported lab from EVE-NG, but including the CLAB topology in this lab makes it so much more portable.
I've created a simple to use docker-compose service to start the lab. ContainerLab it's self is pulled down as a docker container, so you don't need a local installation of ContainerLab unless you want to have one. I recommend you deploy this lab in a Linux environment, as docker virtualization will prevent you from starting in a Mac or Windows env.
Requirements
- Docker
- Docker-Compose
- Roughly 32Gb of available RAM, give or take a few depending on usage.
Starting Lab
Simply run the service using docker-compose
docker-compose run clab
You should see something similar to this
➜ spauto_devnet git:(containerlab) ✗ docker-compose run clab
Creating spauto_devnet_clab_run ... done
INFO[0000] Parsing & checking topology file: spauto-mpls-sdn.yml
INFO[0000] Creating lab directory: /src/clab-spauto-mpls-sdn.yml
INFO[0000] Creating docker network: Name='spauto', IPv4Subnet='172.100.100.0/24', IPv6Subnet='2001:172:100:100::/80', MTU='1500'
INFO[0000] Creating container: xrv-pe1
INFO[0000] Creating container: xrv-rr-1
INFO[0000] Creating container: xrv-p2
INFO[0000] Creating container: xrv-p1
INFO[0000] Creating container: xrv-rr-2
INFO[0000] Creating container: xrv-pe2
INFO[0000] Creating container: xrv-pe3
INFO[0000] Creating container: xrv-pe4
INFO[0001] Creating virtual wire: xrv-pe1:eth4 <--> xrv-pe2:eth3
INFO[0002] Creating virtual wire: xrv-pe2:eth4 <--> xrv-p2:eth1
INFO[0002] Creating virtual wire: xrv-pe4:eth1 <--> xrv-p2:eth6
INFO[0002] Creating virtual wire: xrv-pe1:eth5 <--> xrv-p1:eth2
INFO[0002] Creating virtual wire: xrv-p1:eth5 <--> xrv-p2:eth4
INFO[0002] Creating virtual wire: xrv-p1:eth4 <--> xrv-p2:eth3
INFO[0002] Creating virtual wire: xrv-rr-2:eth1 <--> xrv-p1:eth6
INFO[0002] Creating virtual wire: xrv-rr-2:eth3 <--> xrv-p2:eth5
INFO[0003] Creating virtual wire: xrv-pe3:eth3 <--> xrv-pe4:eth2
INFO[0003] Creating virtual wire: xrv-pe3:eth2 <--> xrv-p1:eth7
INFO[0003] Creating virtual wire: xrv-rr-1:eth4 <--> xrv-p2:eth2
INFO[0003] Creating virtual wire: xrv-rr-1:eth2 <--> xrv-p1:eth3
INFO[0003] Creating virtual wire: xrv-rr-1:eth3 <--> xrv-rr-2:eth2
INFO[0004] Adding containerlab host entries to /etc/hosts file
Run 'containerlab version upgrade' to upgrade or go check other installation options at https://containerlab.srlinux.dev/install/
+---+-----------------------------------+--------------+---------------------------+--------+---------+--------------------+--------------------------+
| # | Name | Container ID | Image | Kind | State | IPv4 Address | IPv6 Address |
+---+-----------------------------------+--------------+---------------------------+--------+---------+--------------------+--------------------------+
| 1 | clab-spauto-mpls-sdn.yml-xrv-p1 | eef17ad66ff1 | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.101/24 | 2001:172:100:100::101/80 |
| 2 | clab-spauto-mpls-sdn.yml-xrv-p2 | 24e5fc6af015 | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.102/24 | 2001:172:100:100::102/80 |
| 3 | clab-spauto-mpls-sdn.yml-xrv-pe1 | efb11d90a5eb | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.11/24 | 2001:172:100:100::11/80 |
| 4 | clab-spauto-mpls-sdn.yml-xrv-pe2 | 7d604f4f06f8 | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.22/24 | 2001:172:100:100::22/80 |
| 5 | clab-spauto-mpls-sdn.yml-xrv-pe3 | 285733f28608 | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.33/24 | 2001:172:100:100::33/80 |
| 6 | clab-spauto-mpls-sdn.yml-xrv-pe4 | 5af4945fab68 | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.44/24 | 2001:172:100:100::44/80 |
| 7 | clab-spauto-mpls-sdn.yml-xrv-rr-1 | b5084e97738d | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.201/24 | 2001:172:100:100::201/80 |
| 8 | clab-spauto-mpls-sdn.yml-xrv-rr-2 | 6a80d3285c2e | h4ndzdatm0ld/vr-xrv:6.1.3 | vr-xrv | running | 172.100.100.202/24 | 2001:172:100:100::202/80 |
+---+-----------------------------------+--------------+---------------------------+--------+---------+--------------------+--------------------------+
This will start a lab topology for you and all the devices should be accessible via SSH/NETCONF/gNMI. The topology YAML file has been crafted with a custom Docker Management Network to statically assign the IP's to each node as well. This network is reserved under 172.100.100.0/24. From your local machine, after starting the devices, you should be able to ssh into each one.
You will know the lab has been successfully started once all devices show healthy state. Validate via docker command
docker ps
The status should say (healthy)
Load Configurations
Unfortunately, these vrnetlab based nodes don't support providing a startup-config via ContainerLab. However, I've provided a Nornir job to simply deploy all the configs to each one of the devices.
After starting the lab and waiting around 7 Minutes, launch off the Nornir script to deploy the configurations.
Install the project venv
poetry install
Activate venv
poetry shell
Locate the Nornir playground
cd mpls_in_the_sdn_era/mpls_sdn_era_nornir
Automate
python nr_deploy_configs.py
Destroying the lab
To destroy the lab, simply override the docker-compose service command and destroy it.
docker-compose run clab containerlab destroy -t spauto-mpls-sdn.yml
Example stats of running all 8 core XR Routers simultaneously.
CONTAINER ID NAME CPU % MEM USAGE / LIMIT MEM % NET I/O BLOCK I/O PIDS
9d6acb79bcbe clab-spauto-mpls-sdn.yml-xrv-pe4 2.56% 1.769GiB / 62.71GiB 2.82% 259kB / 207kB 0B / 82.8MB 13
de723690af86 clab-spauto-mpls-sdn.yml-xrv-p1 2.42% 1.719GiB / 62.71GiB 2.74% 204kB / 212kB 0B / 60.2MB 13
f0e0710d1edb clab-spauto-mpls-sdn.yml-xrv-pe1 2.41% 1.793GiB / 62.71GiB 2.86% 213kB / 168kB 0B / 59.2MB 13
aca6fb91120c clab-spauto-mpls-sdn.yml-xrv-pe3 1.89% 1.758GiB / 62.71GiB 2.80% 121kB / 56.9kB 0B / 52MB 13
63277aac88ef clab-spauto-mpls-sdn.yml-xrv-p2 1.47% 1.718GiB / 62.71GiB 2.74% 198kB / 204kB 0B / 59.1MB 13
c83b2faa5ea8 clab-spauto-mpls-sdn.yml-xrv-pe2 2.02% 1.796GiB / 62.71GiB 2.86% 206kB / 146kB 0B / 57.7MB 13
35fcceedc92a clab-spauto-mpls-sdn.yml-xrv-rr-2 4.33% 1.744GiB / 62.71GiB 2.78% 185kB / 186kB 0B / 60.5MB 13
3f62259d2ae9 clab-spauto-mpls-sdn.yml-xrv-rr-1 1.45% 1.745GiB / 62.71GiB 2.78% 190kB / 190kB 0B / 59.5MB 13
NSO
A pre-packaged NSO Docker container is available. This will be a fresh docker container with the following NEDs:
cisco-ios-cli-6.69
cisco-iosxr-cli-7.33
To spin up the NSO instance
docker-compose up -d nso
To access via UI (admin/admin) get the IP address of the NSO instance and open it in your browser
docker inspect --format='{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}'
If you want to simply ssh from the same machine that's hosting the container
ssh [email protected] -p 2024
The docker-compose service is mapping port 2024 to 22. More to come!