Articles
On this page

Using OpenTSDB in Grafana

The newest release of Grafana adds additional functionality when using an OpenTSDB Data source.

  1. Open the side menu by clicking the the Grafana icon in the top header.
  2. In the side menu under the Dashboards link you should find a link named Data Sources.

    NOTE: If this link is missing in the side menu it means that your current user does not have the Admin role for the current organization.

  3. Click the Add new link in the top header.

  4. Select OpenTSDB from the dropdown.

Name Description
Name The data source name, important that this is the same as in Grafana v1.x if you plan to import old dashboards.
Default Default data source means that it will be pre-selected for new panels.
Url The http protocol, ip and port of you opentsdb server (default port is usually 4242)
Access Proxy = access via Grafana backend, Direct = access directly from browser.
Version Version = opentsdb version, either <=2.1 or 2.2
Resolution Metrics from opentsdb may have datapoints with either second or millisecond resolution.

Query editor

Open a graph in edit mode by click the title. Query editor will differ if the datasource has version <=2.1 or = 2.2. In the former version, only tags can be used to query opentsdb. But in the latter version, filters as well as tags can be used to query opentsdb. Fill Policy is also introduced in opentsdb 2.2.

Note: While using Opentsdb 2.2 datasource, make sure you use either Filters or Tags as they are mutually exclusive. If used together, might give you weird results.

Auto complete suggestions

As soon as you start typing metric names, tag names and tag values , you should see highlighted auto complete suggestions for them.

Note: This is required for the OpenTSDB suggest api to work.

Templating queries

Grafana’s OpenTSDB data source now supports template variable values queries. This means you can create template variables that fetch the values from OpenTSDB (for example metric names, tag names, or tag values). The query editor is also enhanced to limiting tags by metric.

When using OpenTSDB with a template variable of query type you can use following syntax for lookup.

metrics(prefix)               // returns metric names with specific prefix (can be empty)
tag_names(cpu)                // return tag names (i.e. keys) for a specific cpu metric
tag_values(cpu, hostname)     // return tag values for metric cpu and tag key hostname
suggest_tagk(prefix)          // return tag names (i.e. keys) for all metrics with specific prefix (can be empty)
suggest_tagv(prefix)          // return tag values for all metrics with specific prefix (can be empty)

If you do not see template variables being populated in Preview of values section, you need to enable tsd.core.meta.enable_realtime_ts in the OpenTSDB server settings. Also, to populate metadata of the existing time series data in OpenTSDB, you need to run tsdb uid metasync on the OpenTSDB server.

Nested Templating

One template variable can be used to filter tag values for another template varible. Very importantly, the order of the parameters matter in tag_values function. First parameter is the metric name, second parameter is the tag key for which you need to find tag values, and after that all other dependent template variables. Some examples are mentioned below to make nested template queries work successfully.

tag_values(cpu, hostname, env=$env)                   // return tag values for cpu metric, selected env tag value and tag key hostname
tag_values(cpu, hostanme, env=$env, region=$region)   // return tag values for cpu metric, selected env tag value, selected region tag value and tag key hostname

Note: This is required for the OpenTSDB lookup api to work.

For details on opentsdb metric queries checkout the official OpenTSDB documentation