Introducing Cumulus NetQ
Cumulus NetQ is a fabric-wide, telemetry-based validation system, that enables organizations to validate network state, both during regular operations and for post-mortem diagnostic analysis. Running on Cumulus Linux switches and other certified systems — such as Ubuntu, Red Hat, and CentOS hosts — NetQ captures network data and other state information in real time, allowing cloud architects and network operations teams to achieve end-to-end actionable insight from the host to the switch.
The system uses a three-pronged approach to validating networks:
NetQ easily validates potential network configuration changes in a virtualized environment or lab using check, show and trace algorithms, eliminating the need to check nodes one by one and reducing manual errors before they are rolled into production (one of the main causes of network downtime).
NetQ detects faulty network states that can result in packet loss or connectivity issues, and alerts the user with precise fault location data to allow for faster remediation, greatly improving network agility, and reducing downtime costs.
NetQ provides the ability to trace network paths, replay the network state at a time in the past, review fabric-wide event changelogs and diagnose the root cause of state deviations.
What's New in Cumulus NetQ 1.3.0
Cumulus NetQ 1.3.0 includes the following new features and updates:
- The ability to monitor Kubernetes container clusters
- A command line interface for configuring NetQ Notifier
- The ability to decommission NetQ Agents
For further information regarding bug fixes and known issues present in this release, refer to the release notes.
Compatibility with Cumulus Linux
Cumulus NetQ 1.3.0 is compatible with Cumulus Linux versions 3.3.0 through 3.5.z.
NetQ comprises the following components:
- NetQ Agent
- NetQ Telemetry Server
- NetQ Analysis Engine
- NetQ Service Console
Each is described below.
The back-end Python agent installed on every monitored node in the network — including Cumulus Linux switches, Linux bare-metal hosts and virtual machines, or Docker containers. The agent pushes out data to the NetQ Telemetry Server periodically, and when specific
netlink events occur. The agent monitors the following objects via
- address (IPv4 and IPv6)
- route (IPv4 and IPv6)
- bridge fdb
- IP neighbor (IPv4 and IPv6)
Further, every 15 seconds, it gathers data for the following protocols:
- Bridging protocols (LLDP, STP, MLAG)
- Routing protocols (BGP, OSPF)
- Network virtualization (LNV, VXLAN data plane)
- Docker containers
It also listens to the Docker event stream to monitor Docker containers running on a host and gathers container networking information such as NAT translations, networks and container IP and MAC addresses.
NetQ Telemetry Server
The database/key-value store where all network information sent from NetQ Agents running on the network is collected, aggregated and queried from.
NetQ Analysis Engine
The NetQ Analysis Engine is the backend engine utilized when querying NetQ via the CLI, service console, or notifier. The engine has two parts:
The NetQ Agent Command Line Interface. The NetQ CLI can be used on every node and can be used on the NetQ Telemetry Server through
The NetQ command line interface runs on x86 and ARM switches and hosts only.
The NetQ Notifier. The notifier runs on the telemetry server. It responds to events pushed by the NetQ Agent, sending alerts to a configured channel, such as Slack, PagerDuty or
NetQ Service Console
The Service Console provides a browser-based window for accessing the NetQ CLI from anywhere.
Download the Cumulus NetQ User Guide
The Cumulus NetQ user guide is available as a PDF for offline viewing.