Monitoring: SNMP Support

Simple Network Management Protocol (SNMP) is an Internet-standard protocol for collecting and organizing information about managed devices on IP networks, and is used primarily for monitoring network services. SNMP exposes management data in the form of variables (metrics) which describe the system status and configuration. These metrics can then be remotely queried (and, in some circumstances, manipulated) by managing applications.

In RavenDB 3.5 we added support for SNMP which allows monitoring tools like Zabbix, PRTG and Datadog direct access to the internal details of RavenDB. We expose a long list of metrics: CPU and memory usage, server total requests, the loaded databases... And also database specific metrics like the number of indexed items per second, document writes per second, storage space each database takes and so on. The full list can be found here.

You can still monitor what is going on with RavenDB directly from the studio, or by using one of our monitoring tools, but using SNMP might be easier in some cases. As users start running large numbers of RavenDB instances, it becomes unpractical to deal with each of them individually and using a monitoring system that can watch many servers is preferable.

Enabling SNMP in RavenDB

RavenDB 3.5 is already configured to support SNMP and all you have to do is enable it and restart the server. This is done by adding the following key to your app.config/web.config in the appSettings section:

<configuration>
    <appSettings>
        <add key="Raven/Monitoring/Snmp/Enabled" value="true" />

There are two configurable SNMP properties in RavenDB: the SNMP port and the community string.
The default community string is "ravendb" and the default port is 161. You can change those with the following configuration keys:

<add key="Raven/Monitoring/Snmp/Port" value="12345" />
<add key="Raven/Monitoring/Snmp/Community" value="YourString" />

The community string is used like a password. It is sent with each SNMP Get request and allows or denies access to the monitored device.

The Metrics

It is usually easy to query the exposed metrics using a monitoring tool. (Example).
However, you should also be able to access those directly using any SNMP agent like Net-SNMP.

The most basic SNMP commands are snmpget, snmpset and snmptrapd (Read more).

Each metric has a unique identifier (OID) and can be accessed individually.
For example, using the SNMP agent you could run the following snmpget command which gets the server up time metric:

snmpget -v 2c -c ravendb live-test.ravendb.net 1.3.6.1.4.1.45751.1.1.1.2
snmpget -v 3 -l authNoPriv -u ravendb -a SHA -A ravendb live-test.ravendb.net 1.3.6.1.4.1.45751.1.1.1.2

Where "ravendb" is the community string and "live-test.ravendb.net" is the host.

efrat@efrat-VirtualBox:~/SNMP$ snmpget -v 2c -c ravendb live-test-.ravendb.net 1.3.6.1.4.1.45751.1.1.1.2
iso.3.6.1.4.1.45751.1.1.1.2 = Timeticks: (333684) 0:55:36.84

For your convenience we've also added the list of metrics and their associated OIDs here:

RavenDB's root OID id: 1.3.6.1.4.1.45751.1.1.

Templates

For easier setup, we have prepared a few templates for monitoring tools which can be found here.
These templates include the metrics and their associated OIDs.