|

[Aruba DevNet Lecture 5] Completing the NetOps Pipeline with Git and CI/CD

⏱️ Estimated reading time: Approx. 6 minutes (6 min read)

So far we have Python과 Ansible, and New Central APIArmed with this, we've conquered the core elements of network automation one by one. However, in a practical environment, simply having a few script files on your personal computer isn't enough to achieve true automation.

Today, the final installment in the series, we'll explore the "big picture" of how code integrates with a company's business strategy (ITSM) and, combined with cutting-edge AI technologies, transforms into a self-evolving network (AIOps).


1. Why ITIL 5 and CI/CD are necessary for network operations.

In an enterprise environment, the most important things are ‘stability‘ and ‘traceability’.
Just as developers manage source code, so too does the infrastructure environment. IaC (Infrastructure as Code)should be managed as follows.

  • Version Control (Git): 3rd lectureUpload the switch VLAN configuration code created in to Git. This provides complete traceability of who changed the network configuration, when, and why. Now, your code becomes both a blueprint and a resume for your network.
  • CI/CD pipeline integration: When the code is modified (committed), GitLab or GitHub ActionsIt automatically checks the grammar and deploys it to the test network first.

2. [Architecture] CI/CD-based network deployment workflow

The overall flow is exactly the same as the software development process.

  1. Plan & Code: An engineer writes code to add a new VLAN in VS Code.
  2. Commit & Push: Push the changed code to your internal Git repository.
  3. Review & Merge: A senior engineer reviews and approves the code.
  4. CI/CD Pipeline Integration: Once the code is merged, GitHub Actions or GitLab CI will take effect.
  5. Deploy (CD): In an isolated environment (Runner) within the pipeline ansible-playbook The command is automatically executed and reflected on the Aruba CX switch.

3. [Practical Example] Automating Ansible Deployments with GitHub Actions

A simple pipeline configuration file that automatically deploys the Ansible playbook written in Lesson 3 through GitHub Actions (.github/workflows/deploy_network.yml) is an example.

YAML
name: Deploy Aruba CX Network Config

This pipeline will only run when code is pushed to the # 'main' branch.
on:
  push:
    branches:
      - main

jobs:
  deploy-to-switches:
    runs-on: ubuntu-latest

    steps:
      # 1. Import the code from the repository into the Runner.
      - name: Checkout Code
        uses: actions/checkout@v3

      # 2. Install the ensemble in the pipeline environment.
      - name: Install Ansible
        run: |
          sudo apt update
          sudo apt install -y ansible

      # 3. Install Ansible Collection for Aruba CX switch control.
      - name: Install Aruba CX Collection
        run: ansible-galaxy collection install arubanetworks.aos_cx

      # 4. Deploy the configuration to the actual switch. (For security, use GitHub Secrets for the password.)
      - name: Run Ansible Playbook
        env:
          ANSIBLE_HOST_KEY_CHECKING: ""False""
        run: |
          ansible-playbook -i inventory.ini vlan_setup.yml \
          -e "ansible_password=${{ secrets.ARUBA_SWITCH_PASSWORD }}""

Once this pipeline is built, you will no longer need to connect directly to the switch. Just modify the code and push.

4. Evangelist Insights: Harmonizing ITIL 5 and NetOps

Introducing CI/CD to your network isn't just a technical fad.
This is the company's ITSM (IT Service Management) 및 ITIL 5 FrameworkIt is strongly combined with .

The core of ITIL 5 ‘Change Enablement’ Think about the process.

A Git "Pull Request" becomes a change approval request, and the process of deploying approved code through CI/CD is itself a complete change control process. What if a failure occurs? Simply pressing the "Revert" button in Git instantly rolls back to the previous network state.

This is the true way to achieve business stability and IT agility at the same time. Automation Ecosystemno see.


Concluding the DevNet series...

Automation doesn't happen overnight.

Just as the small API call tests in Lecture 1 combined to create the massive CI/CD pipeline of Lecture 5 today, I hope you can start with a small script in your own field.

HPE Aruba Networking's powerful APIs and ecosystem will be your unwavering weapon.

Similar Posts

  • [Aruba DevNet Lecture 4] Central Webhooks Meet Slack

    ⏱️ Estimated reading time: Approx. 3 minutes (3 min read) A network engineer's destiny is a 'battle against failures.' However, if you only become aware of a failure after receiving a call in the middle of the night, it is already too late. Today, utilizing Aruba Central's Webhook feature, we will discuss intelligent monitoring that immediately sends notifications to Slack when specific events (AP Down, switch port errors, etc.) occur…

  • Role-based zero trust environments

    ⏱️ Estimated reading time: Approx. 8 minutes (8 min read) Until now, our network security has been very static. We controlled traffic and strengthened security by creating security policies based on source IP addresses, destination IP addresses, and ports. However, times have changed. IP addresses are also shifting to a DHCP environment, and user devices, including personal devices, are being controlled by the IT team…

  • AOS-CX Security Settings Guide (1)

    ⏱️ Estimated reading time: Approx. 11 minutes (11 min read) Awareness of network security is growing. Traditional network security mostly involved configuring IP address-based firewalls. However, as many applications and systems are now connecting to networks, the security of network switches is also becoming increasingly important. As networks become more complex and interconnected…

  • [Advanced Guide] Optical Transceiver Technology Innovations Supporting AI Data Centers (AIDC)

    ⏱️ Estimated read time: Approx. 9 min read As large-scale AI clusters and AIDC (AI Data Center) environments become commonplace, the key bottlenecks in network engineering have shifted beyond the 'software stack' to 'Physical Signal Integrity' and 'Thermal Management.' In fabrics where tens of thousands of accelerators perform All-to-All communication, even delays of a mere few nanoseconds (ns) and link flaps…

  • [ACSA Education #26] RSTP Elements

    ⏱️ Estimated read time: Approx. 6 minutes (6 min read) Continuing from last time, we would like to explore STP (Spanning Tree Protocol), specifically RSTP (Rapid STP). We will examine the key factors that the RSTP algorithm uses to determine which ports to use for traffic delivery and which not. Bridge Identifier Bridge Protocol Data Unit Rapid Spanning Tree…

  • [ACSA Training #29] Rate Limit & Loop Protection

    ⏱️ Estimated read time: Approx. 2 minutes (2 min read) Rate Limit If the Spanning Tree configuration is incorrect, a Layer 2 loop may be created. In this situation, AOS-CX specifies the amount of traffic that can be received on the Ethernet port or Link Aggregation Group (LAG), such as broadcast, multicast, or unknown unicast traffic…