Optical Network Telemetry with Streaming Mechanisms using Transport API and Kafka

Abstract

We present a streaming mechanism for optical networks based on the Kafka architecture and protocols, to efficiently distribute state and network updates following the upcoming ONF Transport API streaming implementation agreement. The proposed mechanism is validated and experimentally evaluated.

Dynamic Reconfiguration of WDM Virtual Network Topology over SDM Networks for Spatial Channel Failure Recovery with gRPC Telemetry

“…To avoid false alarms, the SXC Degradation analyzer also needs to get in real-time the list of active connections in the WDM VNT SDN controller. In [4], we presented a streaming mechanism for optical networks based on the Kafka architecture and protocols. It enables an efficient distribution of the state and network updates following the ONF Transport API streaming implementation agreement.…”

Architecture to Deploy and Operate a Digital Twin Optical Network

“…To support this critical objective, ONF Transport API has been extended to provide network telemetry information based on information streams [5]. The introduction of a data collector based on publish/subscribe event bus (such as Apache Kafka) provides many benefits that have been presented in [6].…”

“…This paper presents for the first time an architecture to deploy and operate a DTON, and validates it experimentally. Previous work on data streaming [6] has been the keystone for collecting the necessary data for deploying and updating the DTON. To this end, setup and tear-down delays have been measured for a small-scale DTON, based on the ADRENALINE testbed [7].…”

“…Previously to DTON deployment, the physical SDN Controller (typically an Open Line System -OLS-controller) shall subscribe to a data collector (e.g., Kafka broker) in order to publish topological and connectivity data streams. In [6], the authors have demonstrated the usage of ONF TR-548 [5] streaming mechanism that handles the providing of information from one system to another in some form of steady and continuous flow based on an event source/server streaming mechanism.…”

“…The presented DTON life-cycle (Fig.1 right) consists of four phases: a) DTON deployment, where a DTON is requested by the network operator and deployed using virtualization techniques; b) update DTON with real-time adquired data so that DTON reflects current physical network state; c) operate DTON, allowing the network operator to run what-if scenarios and make feasibility studies of novel lightpath deployment; and d) remove DTON.Previously to DTON deployment, the physical SDN Controller (typically an Open Line System -OLS-controller) shall subscribe to a data collector (e.g., Kafka broker) in order to publish topological and connectivity data streams. In[6], the authors have demonstrated the usage of ONF TR-548[5] streaming mechanism that handles the providing of information from one system to another in some form of steady and continuous flow based on an event source/server streaming mechanism.DTON deployment refers to the first stage of DTON life-cycle, where a user requests the deployment of a DTON. This triggers the subscription of SMM to Data Collector, thus all stream-record updates from OLS controller are received (i.e., node and link updates).…”