Unlocking Scalable Network Automation: Lessons from 360’s Ops Strategy

Network Automation

When network devices expand from hundreds to thousands or even tens of thousands, the sheer scale triggers qualitative changes that require new operational models capable of detecting problems and bottlenecks before they impact services.

What Is Network Automation?

Network automation is not limited to writing scripts or relying solely on SDN; it is a broader mindset that embraces any technique that improves efficiency and frees engineers from repetitive tasks.

Automation Types

Script automation – using telnet/SSH libraries for bulk actions such as configuration backup and health checks.

Automated onboarding – employing technologies like ZTP to let devices join the network automatically.

Higher‑level orchestration – abstracting sequences of actions (e.g., integrating network devices with OpenStack) and exposing north‑bound APIs for business services.

Why Network Automation Is Needed

Device counts have grown exponentially while staffing has not kept pace, making traditional manual operations unsustainable and leading to constant firefighting.

Considerations for Network Devices

Devices must meet data‑center redundancy, buffer, and table‑size requirements, and they should support automation‑friendly features such as ZTP, APIs, EVPN, and VXLAN to future‑proof deployments.

Device Auto‑Onboarding

Most major vendors support ZTP. Two common approaches are:

Mapping MAC addresses to configurations in advance and letting devices fetch their config via DHCP.

Combining ZTP with LLDP so a device discovers its location and retrieves the appropriate configuration without pre‑registered MAC mappings.

Business Integration

360’s network team integrates automated onboarding with OpenStack, initially using VLANs and later adopting VXLAN/EVPN for flexible virtual machine migration. They provide a Neutron plugin and also expose generic APIs so other platforms can invoke network actions without being tightly coupled to business code.

Fault Diagnosis

Rapid traffic spikes are identified by correlating switch ports, MAC addresses, and server inventories via CMDB integration, reducing manual lookup effort.

Backbone Monitoring

Distributed ping monitoring across IDC and carrier networks provides proactive health checks and alerts, enabling issues to be spotted before they affect services.

Host‑to‑Host Diagnosis

A path‑query tool allows operators to input a business IP and instantly retrieve the network path, automatically diagnosing common issues such as port errors or high traffic, dramatically cutting troubleshooting time.

Conclusion

The 360 network team’s mission is to provide stable, visible services and free engineers from repetitive work, enabling creative tasks. Their automation efforts—spanning scripting, ZTP, orchestration, and API exposure—have delivered measurable progress, and they will continue to invest in NetDevOps to meet evolving business needs.