Home/ Products/ Automation & Programmability
OcNOS 7.0  ·  Ansible  ·  NETCONF/YANG  ·  gNMI  ·  ZTP  ·  Containers

Automation &
Programmability

Your existing NetDevOps tools work on day one. OcNOS exposes every configuration and operational state through standard, open interfaces, with no screen-scraping and no proprietary orchestrators.

Book a Demo Try OcNOS Free → Read the Docs →
Network teams are shifting from CLI-first operations to programmable, API-driven automation. OcNOS is built for this transition. Every protocol state, configuration element, and operational metric is accessible through Ansible, NETCONF/YANG, gNMI, and OpenConfig: the same standard, model-driven interfaces Cisco IOS-XR and Juniper Junos expose. The difference is the platform: open hardware, open software, and a fraction of the cost.
Application Note

The playbooks engineers use.

A field-proven walkthrough of integrating Ansible with OcNOS: modules, inventory patterns, and the Day-0/1/2 playbook structure.

App Note

Automating Network Management: Integrating Ansible with OcNOS

Official IP Infusion Ansible Collection: modules, inventory patterns, and playbook examples covering BGP, EVPN-VXLAN, and OcNOS-specific operations.

Get the app note
The problem OcNOS automation solves
Before

Manual CLI operations

Config pushed device-by-device over SSH. SNMP polling with 5-minute intervals. Engineers on-site to rack and cable new hardware. No structured state, no rollback.

→
After OcNOS

Fully programmable, Day 0 to Day N

Ansible deploys BGP and EVPN-VXLAN across 100 switches in minutes. gNMI streams telemetry with sub-second updates. ZTP boots new hardware with zero human touch. Rollback in seconds.

Programmable Interfaces

Six ways to automate OcNOS

Pick one or use all six together. OcNOS doesn't force a proprietary orchestrator. It works with the tools your team already uses.

Ansible Collection

ansible-galaxy collection install ipinfusion.ocnos

Official IP Infusion collection on Ansible Galaxy. Modules include ocnos_facts, ocnos_config, ocnos_command, ocnos_bgp_facts, and ocnos_isis_facts. Jinja2 templating for dynamic, fleet-wide provisioning. Use the same Ansible playbooks you already write. OcNOS speaks the same language as Arista, Juniper, and Cisco.

ansible-playbook deploy-sr-mpls.yml
$ansible-playbook -i inventory deploy-sr-mpls.yml
ok: [spine-01] - IS-IS + SR-MPLS configured
ok: [leaf-01..04] - BGP-LU overlay up
ok: [all] - config backup saved
PLAY RECAP · 5 switches · 0 failures · 0:42 elapsed
Ansible Guide in OcNOS Docs

NETCONF / YANG 1.1

IPI-native + OpenConfig models · RFC 6241

Full NETCONF 1.1 with both IPI-native and OpenConfig YANG data models. Structured config and operational state retrieval: get, edit-config, commit, rollback, and candidate datastore. Works with Python's ncclient library, Ansible's netconf_config module, or any NETCONF client.

netconf: edit-config + commit
<edit-config>
<target><candidate/></target>
<config>...YANG payload...</config>
</edit-config>
<ok/> ← candidate locked
<commit/> ← applied atomically

gNMI Streaming Telemetry

OcNOS 7.0 · Dial-in + Dial-out · On-Change + Periodic · TLS

Real-time, push-based telemetry via gNMI over gRPC. Replaces SNMP polling with structured, sub-second updates across interfaces, BGP session state, MPLS forwarding, QoS queues, PFC/ECN counters, and system health. Connect to Telegraf, Prometheus, or any gRPC collector, then visualise in Grafana.

gnmic subscribe: interfaces
$gnmic subscribe --path /interfaces/interface[name=eth1]/
in-octets: 4,218,753,012
out-octets:3,874,112,440
mode: ON_CHANGE · encoding: PROTO

Zero Touch Provisioning

DHCP + TFTP/HTTP · Day 0 automation

Rack it. Cable it. Power it on. ZTP handles everything else. DHCP assigns an address and points to a provisioning server. The switch downloads its OS image and startup config over TFTP or HTTP, applies it, and joins the network, with no console cable and no engineer on-site. Combine ZTP with Ansible for a complete Day 0-to-Day 2 pipeline.

Real-world use case: A 48-node AI GPU cluster spine-leaf fabric provisioned over a weekend with zero on-site engineers. ZTP boots each switch, Ansible deploys BGP + PFC configuration, gNMI confirms lossless fabric.

IP Maestro EMS

Web UI · REST API · OcNOS element management

For teams that want a GUI alongside API automation. IP Maestro is a GUI-based Element Management System (EMS) for OcNOS: an interactive topology map, inventory and device details, fault and performance monitoring, configuration and software management, and streaming-telemetry visualization. It talks NETCONF to OcNOS devices and exposes a northbound REST API so a higher-level controller or OSS can integrate with it.

Explore IP Maestro

On-Switch Containers

OcNOS 7.0 · Container runtime (K3S) · single container

OcNOS 7.0 includes an on-switch container runtime (K3S) with single-container lifecycle management. Run a Telegraf agent, a Zabbix proxy, a custom Python or Go script, or a security tool directly on the switch, alongside the NOS, without external compute hardware. Standard SSH, SCP, SFTP, and FTP move images and configs.

Also: OcNOS exposes standard SNMP MIBs, so it can be monitored with Zabbix's generic SNMP templates, keeping legacy SNMP-based NOC tooling working alongside gNMI.

Cloud-native operations

Model-driven, transactional, and resilient by design

The operating model engineers expect from a modern NOS: structured state, safe changes with rollback, streaming visibility, and control-plane resilience. Standard interfaces, open hardware.

Model-driven configuration

OpenConfig + IPI-native YANG · NETCONF 1.1

Configuration and operational state are structured data, not screen-scraped text. Edit the candidate datastore, then commit atomically or roll back, the same model-driven workflow Cisco IOS-XR and Juniper Junos use, over standard NETCONF 1.1 with OpenConfig and IPI-native YANG models.

Streaming telemetry, both directions

gNMI · dial-in + dial-out · TLS · multi-VRF

gNMI Subscribe (dial-in) and Publish (dial-out over grpctunnel), in on-change and periodic modes, secured with TLS and authenticated per user. Runs in-band across a single or multiple VRFs on OpenConfig models. Push-based, structured state for Telegraf, Prometheus, and Grafana, in place of SNMP polling.

Control-plane resilience

Graceful Restart / NSF · BGP · OSPF · IS-IS

Graceful Restart with Non-Stop Forwarding keeps the data plane forwarding while the control plane restarts, for BGP, OSPF, OSPFv3, and IS-IS. Process restarts and planned maintenance do not drop transit traffic.

On-switch programmability

Container runtime (K3S) · ZTP · SSH/SCP/SFTP

Run an agent or tool in a container directly on the switch through the built-in runtime (K3S, single-container lifecycle), so a collector, a script, or a probe runs alongside the NOS with no external compute. ZTP boots new hardware hands-free, and SSH, SCP, SFTP, and FTP move images and configs.

gNMI Streaming, Not Polling
6 Automation Interfaces
0 Proprietary Orchestrators Needed
600+ Production Networks
Day 0 First Automated Push
Full Lifecycle Coverage

Day 0 → Day 1 → Day 2

OcNOS covers the complete operational lifecycle with open, standard tooling. No proprietary orchestrators, no CLI scripting, no SNMP polling.

Day 0: Provision

Hardware joins the network itself

Switch powers on. DHCP assigns an IP. ZTP downloads the OcNOS image and base config over TFTP/HTTP. The device comes up ready to manage. No console cable. No engineer on-site. Entire rack provisioned while your team sleeps.

ZTP DHCP OcNOS
Day 1: Configure

Services deployed in minutes, not days

Ansible playbooks push BGP, MPLS, EVPN-VXLAN, QoS, ACLs, and timing config across the fleet. NETCONF/YANG ensures structured, validated pushes with rollback on failure. IP Maestro, the GUI-based EMS for OcNOS, adds GUI-based config push and scheduling.

Ansible NETCONF/YANG OpenConfig IP Maestro
Day 2: Operate

Real-time visibility, automated remediation

gNMI streams telemetry to Telegraf → Prometheus → Grafana. Ansible handles config drift detection, scheduled backups, and software upgrades. An on-switch container runs custom tools. IP Maestro provides topology, fault monitoring, and scheduled config and software updates.

gNMI Telegraf Grafana Containers Zabbix
Ecosystem

Works with your existing stack

OcNOS speaks standard protocols. No forklift of your toolchain. The tools your team already knows work on day one.

Config Management

Ansible

Official Galaxy collection: ipinfusion.ocnos

Telemetry Collector

Telegraf

gNMI input plugin, direct subscription

Metrics Store

Prometheus

gNMI exporter, time-series metrics

Visualisation

Grafana

Pre-built dashboards for OcNOS telemetry

SNMP Monitoring

Zabbix

Standard SNMP MIBs, works with Zabbix SNMP templates

Python NETCONF

ncclient

Standard Python library, full NETCONF 1.1

gNMI CLI Client

gnmic

Subscribe, get, set: command-line testing

On-Switch Compute

K3S runtime

Single-container runtime, native in OcNOS 7.0

Common Questions

OcNOS automation, answered

Questions network engineers ask before deploying.

Does OcNOS support Ansible?
Yes. IP Infusion publishes an official Ansible Collection on Ansible Galaxy (ipinfusion.ocnos). It includes modules for facts gathering, command execution, configuration management, and protocol-specific operations including ocnos_facts, ocnos_config, ocnos_command, ocnos_ping, ocnos_bgp_facts, and ocnos_isis_facts. Playbooks use Jinja2 templating for fleet-wide, dynamic provisioning across hundreds of OcNOS devices.
What YANG models does OcNOS support?
OcNOS supports both IPI-native YANG models and OpenConfig models over NETCONF 1.1. This allows structured configuration and operational state retrieval: get, edit-config, commit, rollback, and candidate datastore operations. The same model-driven approach used by Cisco IOS-XR and Juniper Junos. Python's ncclient library works directly with OcNOS NETCONF.
Can I stream OcNOS telemetry to Grafana or Prometheus?
Yes. OcNOS 7.0 supports gNMI streaming telemetry with both on-change and periodic subscription modes, in dial-in and dial-out (grpctunnel) styles, secured with TLS and per-user authentication. Connect directly to Telegraf (gNMI input plugin), Prometheus (gNMI exporter), or any gRPC-compatible collector and visualise in Grafana. Sensor paths cover interfaces, BGP state, MPLS labels, QoS queues, PFC counters, and system health, replacing SNMP polling with push-based, structured data.
What is Zero Touch Provisioning (ZTP) in OcNOS?
ZTP automates Day 0 device setup entirely. When a new switch powers on, it uses DHCP to locate a provisioning server, then downloads the correct OcNOS image and startup configuration via TFTP or HTTP, with no console cable and no engineer on-site. Combined with Ansible, ZTP covers the full Day 0-to-Day 2 lifecycle: ZTP handles initial boot and image delivery, Ansible handles ongoing config management, and gNMI handles operational monitoring.
Can I run containers directly on OcNOS?
OcNOS 7.0 includes an on-switch container runtime (K3S) for single-container lifecycle management. Run a third-party monitoring agent such as Telegraf or a Zabbix proxy, a security tool, or a custom Python or Go script directly on the switch, alongside the NOS, without external compute. Standard SSH, SCP, SFTP, and FTP handle image and configuration transfer.
Is OcNOS cloud-native?
OcNOS is a model-driven, API-first NOS built for automated operations on open hardware. Configuration and operational state are structured data through OpenConfig and IPI-native YANG over NETCONF 1.1, with a candidate datastore, atomic commit, and rollback rather than screen-scraped CLI. Telemetry streams over gNMI (dial-in and dial-out, on-change and periodic, TLS-secured, single or multi-VRF) instead of SNMP polling. Graceful Restart with Non-Stop Forwarding keeps the data plane forwarding through control-plane restarts, and a built-in container runtime hosts agents on the switch. Every interface is standard, so the same NetDevOps pipeline that manages Cisco or Arista manages OcNOS, on Broadcom whitebox hardware at lower cost.
Does OcNOS work with Zabbix for SNMP monitoring?
Yes. OcNOS exposes standard SNMP MIBs, so it can be monitored with Zabbix using Zabbix's generic SNMP device templates. For teams still using SNMP-based NOC tooling, OcNOS runs SNMP alongside gNMI. You can operate both in parallel during a migration to streaming telemetry.
How does OcNOS automation compare to Arista EOS or Cisco NX-OS?
OcNOS exposes the same standard interfaces, NETCONF/YANG, gNMI, OpenConfig, and Ansible, that Arista EOS and Cisco NX-OS do. The difference is platform: OcNOS runs on validated open hardware from Edgecore, UfiSpace, and other whitebox vendors, at significantly lower total cost. Your existing Ansible playbooks and Grafana dashboards can migrate to OcNOS with minimal changes.
Automate These Products

One toolchain for your entire open network

Every automation interface on this page works with OcNOS-SP, OcNOS-DC, and IP Maestro: one consistent pipeline across your estate.

Service Provider NOS

OcNOS-SP

Automate SR-MPLS, BGP, and mobile backhaul deployments with Ansible. Stream IEEE 1588v2 timing state and MPLS forwarding metrics via gNMI. ZTP for cell site gateway rollouts.

OcNOS-SP product page → Data Center NOS

OcNOS-DC

Deploy EVPN-VXLAN leaf-spine fabric with Ansible in under an hour. Stream PFC and ECN counters via gNMI for AI fabric visibility. ZTP for GPU cluster rack onboarding at scale.

OcNOS-DC product page → OcNOS EMS

IP Maestro

The GUI element management system (EMS) for OcNOS. Topology visualisation, fault and performance monitoring, telemetry charts, config push and scheduling, and software-image updates, with a northbound REST API.

IP Maestro product page →
Get Started Today

See OcNOS automation in action

Book a live demo with our engineering team. Bring your automation requirements, and we'll show you exactly how OcNOS integrates with your pipeline.

Book a Technical Demo Try OcNOS Free → Read the Docs →