SNMP generic server extension

Latest Dynatrace
Extension
Published Feb 19, 2026

Monitor server infrastructure health and performance using SNMP protocol in environments where you can't deploy OneAgent.

Get Started

Overview

This extension collects generally supported SNMP infrastructure metrics to monitor the health and resource usage of servers. Metrics are collected through SNMP get polling.

Use cases

Monitoring of server infrastructure where a OneAgent isn't feasible to install.
Unix Servers: As an alternative to: Remote Unix
Windows Servers: As an alternative to WMI: Remote Windows

Compatibility information

Any device which supports SNMP polling and one or more of the following MIBs:
- HOST-RESOURCES-MIB
- UCD-SNMP-MIB
- ENTITY-SENSOR-MIB
Unix: Net-SNMP software agent running. Red Hat documentation example
Windows: the SNMP service needs to be enabled and configured. Microsoft documentation

Activation and setup

Simply activate the extension in your environment using the in-product Hub and provide the required device connection configurations. This extension uses the Dynatrace SNMP Data source.

Details

This extension collects SNMP OID metrics from commonly supported MIBs. Most feature sets map to general entries or tables within the MIB files.

SNMPv2-MIB: System properties such as sysname and the sysuptime metric.
HOST-RESOURCES-MIB: Monitoring basic system resources such as CPU, memory, storage, and running processes on a host.
UCD-SNMP-MIB: Monitoring detailed system metrics such as load averages, disk I/O
ENTITY-SENSOR-MIB: Monitoring of physical sensors such as temperature, voltage, fan speed, and others.

FAQ

What is the Custom Metric Usage of this extension?

The custom metric usage can be managed through which feature sets you choose to enable. sysuptime is the only enforced metric as part of the 'default' feature set. You can also use filter variables to restrict collection of specific paths and sensor types Dimension Filters All other feature sets are based on an Object Entry or Table within the SNMP Mib. e.g., hr-storage fs = HR hrStorageStable. Sometimes this is split into multiple for even more granular selection such as ucd-system-cpu-basic, ucd-system-cpu-detailed and ucd-system-swap which all come from: UCD systemStats.

To estimate the metrics ingested you can use a formula such as:

Scalar OIDs: 1 metric per OID
#metric_keys_in_fs

Table OIDs: 1 metric per each table entry (e.g., sensor, disk, process, file):
#metric_keys_in_fs * #entries_in_table