Open Observability

Hedgehog includes open observability for Zero Touch Lifecycle Management.  Unlike the network industry incumbents, we don't charge you for telemetry, collection or dashboards.  Everything is open-source and available by default with our software appliance. 

We give you dashboards for network metrics, logs and traces with options to run them on Grafana Cloud, Grafana Enterprise, or other time series databases that support the Prometheus API.  With Hedgehog you get fully automated:

  • Dashboards
  • Collection
  • Logs
  • Metrics
  • Traces
Get Hedgehog
hh-icon_observability

Dashboards

Comprehensive Network Visibility with Built-In Grafana Dashboards

Hedgehog includes network observability dashboards out of the box.  Our software appliance includes Grafana dashboard templates for network resources, fabric, interfaces, logs, switch platform, and linux platform metrics.  

grafana_icon

Network Resources

This dashboard reports usage and capacity of switch programmable network resources such as:

  • ACLs
  • IPv4 Routes
  • IPv4 Nexthops
  • IPv4 Neighbors
  • IPMC Table
  • FDB

 

 

hedgehog-asic-critical-resourses-stats

Fabric

A BGP fabric refers to a network topology or architecture that uses the Border Gateway Protocol (BGP) as the primary routing protocol to interconnect switches, routers, or network nodes—often within data centers, service provider networks, or modern cloud infrastructures.  Monitor the VPC fabric underlay and external peering with metrics for:

  • BGP Neighbors
  • BGP Session state
  • Number of BGP Updates and Prefixes sent/received for each BGP Neighbor
  • Keepalive counters
hedgehog-bgp-fabric-stats

Interfaces

A network interface is a point of interaction between a computer (or any networked device) and a network. It enables the device to send and receive data across the network.  Use Hedgehog to monitor these interface metrics:

  • Interface Oper/Admin state
  • Total input/output packets counter
  • Input/output PPS/Bits rate
  • Interface utilization
  • Counters for Unicast/Broadcast/Multicast packets
  • Errors and discards counters
hedgehog-bgp-fabric-stats

Power and Cooling

Power and cooling matter a lot to keep cloud networks up and running.  Use this dashboard to monitor the power supply unit, temperature sensors, and fan trays in your switches:

  • Input/output PSU voltage
  • Fan speed
  • Temperature from switch sensors (CPU, PSU, etc)
  • For transceivers with DOM - optic sensor temperature
hedgehog-power-cooling

Linux Platform

Network operating systems run on Linux and require their own CPU, memory and disk space to in turn provide network resources.  

Grafana Node Exporter Full is an open source Grafana board that monitors Linux resources required by the NOS, including

  • Memory
  • Disk
  • CPU
hedgehog-grafana-node-full

Telemetry Collection

Unified Telemetry Collection with Grafana Alloy Integration

To generate dashboards, you need to collect telemetry data from your network.  The Hedgehog software appliance includes Grafana Alloy with your download.  Grafana Alloy is an open-source telemetry collector developed by Grafana Labs, designed to help unify and simplify the collection of logs, metrics, and traces from various sources.  It is essentially a distribution of the open-source project known as OpenTelemetry Collector, combined with features from other Grafana tools like Promtail, Agent, and integrations for Prometheus, Loki, and Tempo.

You can use Hedgehog to enable enable Grafana Alloy on your switches to forward metrics and logs to the configured targets using Prometheus Remote-Write API and Loki API. Metrics includes port speeds, counters, errors, operational status, transceivers, fans, power supplies, temperature sensors, BGP neighbors, LLDP neighbors, and more. Logs include Hedgehog agent logs.

You can enable Hedgehog observability any time in your Zero Touch Lifecyle.  Like usual, just modify some Hedgehog YAML.  Hedgehog will then push logs and metrics to Grafana Enterprise running in your private cloud, or you can use Hedgehog with Grafana Cloud right out of the box.

alloy_icon

Logs

Efficient Log Aggregation with Grafana Loki Integration

Grafana Loki is a horizontally scalable, highly available, multi-tenant log aggregation system developed by Grafana Labs, designed to store and query logs in a way that is efficient, cost-effective, and deeply integrated with Grafana dashboards.

Unlike traditional log aggregators (e.g., Elasticsearch + Logstash), Loki is "prometheus-inspired", meaning it organizes logs with labels rather than indexing their full contents, making it much more efficient for many observability use cases.

Use Hedgehog out of the box with Grafana Cloud Logs, or use the Hedgehog API to push network logs to your Grafana Enterprise Loki instance.  

loki-icon

Metrics

Scalable Network Metrics Monitoring with Grafana Mimir

Grafana Mimir is a new distributed time series database and monitoring engine developed by Grafana Labs. It is designed to be a long-term, scalable, and cloud-native replacement or complement to Prometheus, with deep integration into Grafana Cloud and modern observability stacks.

While Prometheus is well-established and extremely popular for metrics collection and monitoring, Mimir is built to address some of its limitations, especially around horizontal scalability, global querying, and long-term storage.  

Hedgehog pushes network metrics to Prometheus-compatible APIs including Mimir, Thanos, Grafana Cloud, and Influx DB.  Hedgehog network metrics include Metrics includes port speeds, counters, errors, operational status, transceivers, fans, power supplies, temperature sensors, BGP neighbors, LLDP neighbors, and more.  You can monitor your network metrics in Grafana Cloud Metrics, Grafana Mimir in your Grafana Enterprise instance, or alternatives like Thanos.  

Prometheus_software_logo

Traces

End-to-End Network Traceability for Enhanced Observability

A trace is a record of a single transaction or request as it flows through a system — especially a distributed system or microservices architecture.  A trace is composed of spans, each representing a unit of work (e.g., a function call, API request, DB query, or a network flow).

Each span includes metadata like start and end timestamps, tags/attributes (e.g., service name, error status), parent-child relationships between spans.  Together, these spans provide a complete timeline and dependency tree of what happened during the execution of a request. 

Traces can be particularly useful in AI training where training runs may fail due to large "elephant" network flows that create congestion on the network, degrade GPU utilization and may cause training runs to fail.   

Later this year Hedgehog will add network traces to our open telemetry collection so you can analyze network flows in Grafana Cloud Traces or Grafana Tempo in your Grafana Enterprise stack.  

grafana-tempo