Open Networking Enterprise Suite (ONES)

Introduction and Overview

Open Networking Enterprise Suite (ONES) is a Network Orchestration, Visibility, and Assurance solution for multi-vendor and multi-NOS operated Network Infrastructure. ONES provides a one-stop solution from delivering deep visibility into your datacenter networks to extending 24x7 support functions for SONiC. It also hosts a powerful analytics engine that assists users to identify network issues and troubleshoot their networks, in case of common network anomalies and disruptions.

ONES uses Auto-discovery for SONiC devices and a YAML or CSV-based template for adding non-SONiC devices during the onboarding process and continuously collects streaming telemetry data from them to provide insights on:

• Data Center Inventory

• Network State

• Platform and System Health

• Control and Data Plane resource Utilisation

• Traffic Utilisation

• Software Compliance

• AI Fabric

• Underlay and Overlay protocols view

ONES monitors various control and data plane metrics to provide these insights.

Rule Engine

ONESv3.0 application has the capability to trigger notifications via Slack app notifications when certain user-defined threshold values are breached.

In data centre operations, a rule engine with alerts for various metrics is essential for proactive monitoring and management of critical components and services. Rule Engine pushes the configured rule notification in case any device breaches the threshold value configured under the rule to SLACK Channel, Zendesk Support and ServiceNow Ticketing Service integration

Let's see the different types of rule engine metrics for specific Entity/features in a data centre environment

1. CPU and Memory Utilisation

2. Fan and PSU LED status

3. Traffic Bandwidth

4. ASIC Routes

5. Health Services

6. Traffic Errors and Discard Counters

7. BGP Neighbours flapping notification

8. Device down status

9. Link flap status

10. Device SSD Memory Utilization, Health and Temperature

11. ROCE Counters

ONES Orchestration

ONES orchestration provide network admins to automate the fabric configuration using configuration templates for provisioning physical interfaces, layer 3 configuration for building IP-CLOS fabric using

• BGP as a routing protocol including BGP-unnumbered

• Symmetric/Asymmetric IRB

• BGP Peering with PO

• L2/L3 MC-LAG

• EVPN MultiHoming

• Layer2 Leaf-Spine (L2/L3 Mode)

• Rack-to-Rack Deployment

• BGP Peering over MC-LAG PeerLink

• BGP Peering using separate Link between MC-LAG Peers

• SFLOW

• DHCP Relay

• RoCE Config

• AI Fabric config

• SAG / SVI

• NTP, SNMP, SYSLOG

• Incremental Config update for L2VNI/L3VNI

• Enhanced backup and restore options via UI

• Enhanced API support - Config Replace

ONES orchestration not only configures the fabric but also make sure the Fabric is operational by doing verifying the configuration at every stage.

ONES provides north bound API access for configurations originating from external orchestration tools.

Data Lake

A data lake is a centralized repository that allows you to store vast amounts of structured, semi-structured, and unstructured data in its raw format. Unlike traditional data warehouses where data is stored in a structured manner, a data lake retains the data in its native format until it's needed for analysis or processing. ONES provide the capability to store the RAW data of all the Metrics to Cloud and then user will be able to use that RAW data for any deployment or any other use cases.

Here are key components and characteristics of a ONE DL.

1. Storage of Diverse Data Types: A data lake can store various types of data, including structured data (like relational databases), semi-structured data (like JSON, XML), and unstructured data (like documents, images, videos). This flexibility allows organizations to ingest and store data from different sources without the need for extensive preprocessing.

2. Scalable and Cost-Effective Storage: Data lakes are typically built on scalable storage systems, such as cloud-based object storage (e.g., Amazon S3, Azure Data Lake Storage) or Splunk . These systems can efficiently handle large volumes of data and offer cost-effective storage solutions.

3. Schema-on-Read Approach: In contrast to traditional data warehouses that use a schema-on-write approach (where data must be structured and conform to a predefined schema before storage), data lakes adopt a schema-on-read approach. This means that data is stored in its original form, and the schema is applied at the time of data retrieval or analysis. This flexibility allows users to apply different schemas and interpretations to the same dataset based on their analytical needs.

4. Support for Big Data Processing and Analytics: Data lakes serve as a foundational component for big data analytics and processing. Users can perform various analytics tasks, including exploratory data analysis, data mining, machine learning, and real-time analytics, directly on the data lake. Tools like Apache Spark, Apache Hive, and Presto are commonly used for querying and processing data stored in data lakes.

5. Support for Data Discovery and Self-Service Analytics: Data lakes enable data discovery and self-service analytics, empowering users to explore and analyze data without extensive dependencies on IT teams. Data scientists, analysts, and business users can access relevant data directly from the data lake, speeding up insights generation and decision-making processes.

In summary, ONE DL provides a flexible and scalable platform for storing, managing, and analyzing diverse data types at scale. By leveraging a schema-on-read approach and supporting various analytics tools, ONE DL facilitate advanced data analytics and enable organizations to derive valuable insights from their data assets. However, proper governance, security, and metadata management are crucial to ensure the usability, reliability, and integrity of data lakes.