ONOS (software)
Generated by GPT-5-miniExpansion Funnel Raw 80 → Dedup 0 → NER 0 → Enqueued 0
| ONOS (software) | |
|---|---|
| Name | ONOS |
| Developer | Open Networking Foundation, ON.Lab, Linux Foundation |
| Released | 2014 |
| Programming language | Java (programming language) |
| Operating system | Linux |
| License | Apache License |
ONOS (software) ONOS is an open-source network operating system originally created to advance software-defined networking and network functions virtualization initiatives led by Open Networking Foundation and ON.Lab. It provides a carrier-grade platform intended for service providers, cloud operators, and enterprise network operators seeking programmable control of packet, optical, and wireless infrastructure. The project aligns with trends promoted by Telecom Infra Project, Linux Foundation Networking and interoperates with software such as Open vSwitch, OpenDaylight, KVM (kernel-based virtual machine) and Docker (software).
Overview
ONOS emerged from research and industry efforts involving contributors from Stanford University, Princeton University, California Institute of Technology, Intel Corporation, Broadcom Inc., and major telecommunications vendors. It targets production-grade deployments in contexts similar to projects like OpenStack, Kubernetes, and Apache Hadoop by offering a distributed control plane alternative to monolithic controllers such as Floodlight (software) and Ryu (SDN framework). The platform emphasizes high availability, fault tolerance, and hardware abstraction for devices from vendors including Cisco Systems, Juniper Networks, Huawei, and Nokia.
Architecture
ONOS implements a clustered, distributed control plane where each node runs instances of platform services inspired by distributed systems research from Google, Berkeley Software Distribution, and the Paxos (computer science) family of consensus algorithms. Core architectural components include an intent-based northbound model, southbound protocol adapters for OpenFlow, NETCONF, gNMI, and optical control protocols, and an internal store based on eventual consistency and strong consistency primitives. The design maps to patterns found in Apache Zookeeper, etcd, and Apache Cassandra for state management, while leveraging Apache Karaf and OSGi for modularity similar to what appears in Eclipse (software) projects.
Features and Functionality
ONOS provides intent compilation, topology abstraction, flow programming, device management, path computation, and multi-domain orchestration. It exposes northbound REST and gRPC APIs comparable to interfaces used by OpenStack Neutron, Ansible (software), Terraform (software), and integrations with NETCONF managers. Feature sets include traffic engineering modules inspired by IETF drafts on segment routing, MPLS control interworking similar to RFC 3031, and integration points for BGP route distribution akin to implementations from Quagga and FRRouting. The platform supports application development models paralleling Apache Karaf bundles and OSGi services to allow rapid addition of SDN applications.
Deployment and Scalability
ONOS is designed to scale horizontally across clusters deployed on virtualized infrastructure such as OpenStack clouds, bare-metal farms used by AT&T, and container orchestration platforms like Kubernetes. High-availability capabilities mirror approaches used by Google Spanner and Amazon Web Services for geo-redundancy, while state replication strategies align with practices from Cassandra and etcd. Operators have deployed ONOS in carrier networks, research testbeds including GENI (testbed), and campus environments, managing thousands of devices and millions of flows comparable to large deployments of Facebook and Google network fabrics.
Use Cases and Adoption
Adopters include telecommunications providers, data center operators, and research institutions. ONOS has been applied to use cases such as software-defined wide area networks (SD-WAN), optical transport control, 5G mobile fronthaul/backhaul orchestration, and edge computing integration akin to efforts by Edge Computing Consortium and Mobile Edge Computing (MEC). Notable trials and collaborations have involved entities like Telefonica, NTT, Verizon, China Mobile, and academic testbeds run at University of California, Berkeley and Stanford University. The platform competes and cooperates with projects including OpenDaylight, Tungsten Fabric, and proprietary controllers from Ericsson and Huawei Technologies.
Development and Community
Development of ONOS is coordinated through governance structures characteristic of open-source communities hosted by the Open Networking Foundation and influenced by practices in the Linux Foundation ecosystem. Contributors include engineers from Intel, Broadcom, Ciena, and research groups at Princeton University and University of California, Los Angeles. The project maintains public repositories, issue trackers, and continuous integration pipelines comparable to workflows used by Apache Software Foundation projects and Kubernetes SIGs. Community events and developer summits align with conferences like Open Networking Summit, Interop, and MOBICOM.
Security and Performance considerations
ONOS addresses security and performance through role-based access control integrations similar to solutions from Okta and Keycloak, TLS for southbound and northbound channels like protections specified by IETF TLS standards, and mitigation strategies against distributed failure modes explored in CAP theorem analyses. Performance tuning relies on techniques used in high-performance networking by Cisco Systems and Juniper Networks, including efficient flow rule installation, pipeline offloads with P4 (programming language), and hardware acceleration via DPDK and SmartNIC technologies offered by Mellanox Technologies and Intel. Operational security also requires coordination with standards from ETSI and compliance regimes relevant to service providers.