Key Concepts
Before diving into MetricsHub, familiarize yourself with the core concepts that shape how monitoring is configured and organized.
Resources
A resource is any entity you want to monitor: a server, a network switch, a storage array, a database, or an application. Each resource is identified by:
- A unique resource ID (e.g.,
my-server-01) - A hostname or IP address (
host.name) - A type indicating what kind of system it is (
host.type)
resources:
my-server-01: # Resource ID
attributes:
host.name: 192.168.1.100 # Hostname or IP
host.type: linux # Type of system
Resource Types
| Type | Description | Common Protocols |
|---|---|---|
linux | Linux servers | SSH, SNMP |
win | Windows servers | WMI, WinRM |
network | Switches, routers, firewalls | SNMP |
storage | SAN, NAS, storage arrays | HTTP, WBEM, SNMP |
oob | Out-of-band management (BMC, iLO, iDRAC) | IPMI, HTTP |
aix | IBM AIX systems | SSH |
hpux | HP-UX systems | SSH |
solaris | Oracle Solaris systems | SSH |
Resource Groups
Resource groups are containers that group related resources together. They're useful for:
- Managing large, distributed infrastructures
- Applying shared settings to multiple resources
- Organizing multi-site deployments
For example, resource groups may have a site attribute indicating the physical or logical location (data center, cloud region, etc.):
resourceGroups:
paris:
attributes:
site: paris-dc
resources:
server-1: ...
server-2: ...
london:
attributes:
site: london-dc
resources:
server-3: ...
The site attribute is required for sustainability and hardware monitoring dashboards.
Tip: For small, centralized environments, you don't need resource groups; just use
resources:directly with a top-levelsiteattribute if needed.
Protocols
Protocols define how MetricsHub communicates with your resources to collect metrics. Each protocol has its own authentication and connection parameters.
| Protocol | Use Case |
|---|---|
| SSH | Linux/Unix command execution |
| WMI | Windows local/domain queries |
| WinRM | Windows remote management |
| SNMP v1/v2c | Network devices (community string) |
| SNMP v3 | Network devices (encrypted) |
| HTTP/HTTPS | REST APIs, web services |
| WBEM | VMware, CIM-based systems |
| IPMI | Hardware management (BMC) |
| JDBC | Database queries |
| JMX | Java application monitoring |
A resource can use multiple protocols simultaneously:
resources:
my-server:
attributes:
host.name: server-01
host.type: linux
protocols:
ssh:
username: admin
password: changeme
snmp:
version: v2c
community: public
Connectors
Connectors are YAML-based definitions that describe what to collect and how to interpret the data. They contain:
- Detection rules (how to identify compatible systems)
- Data sources (SNMP OIDs, WMI queries, CLI commands, etc.)
- Metric mappings (how to transform raw data into OpenTelemetry metrics)
MetricsHub includes hundreds of built-in connectors for:
- Operating systems (Linux, Windows, AIX, etc.)
- Hardware vendors (Dell, HPE, Cisco, Lenovo, etc.)
- Storage systems (NetApp, Pure Storage, EMC, etc.)
- Network devices (Cisco, Juniper, Arista, etc.)
See the Connectors Directory for the full list.
Automatic Detection
When MetricsHub connects to a resource, it automatically:
- Tests which connectors are compatible
- Selects the best matching connectors
- Begins collecting metrics
You can override this behavior to force or exclude specific connectors. See Customizing What Gets Monitored.
Monitors
Monitors are the individual components discovered within a resource—such as:
- CPUs, memory, disks
- Network interfaces
- Power supplies, fans, temperatures
- Processes, services
- Storage volumes, LUNs
Each monitor produces metrics that are exported to your observability platform.
Metrics and OpenTelemetry
MetricsHub exports all collected data using the OpenTelemetry standard:
- Metrics follow OpenTelemetry semantic conventions
- Data is exported via OTLP (gRPC or HTTP)
- Compatible with Prometheus, Datadog, Splunk, New Relic, and 30+ platforms
This means your metrics are portable and standardized, regardless of which backend you use.
Get Started
- Choose your edition — Community orEnterprise
- Follow a quick start guide:
- Configure monitoring for your resources