Hi all. I’m very new to this domain, so all thoughts are welcome.
We have a self-hosted TimescaleDB that we’re using to capture metrics and traces via OpenTelemetry. We would like to use Grafana to visualize those traces and metrics. I know that I can use MST to connect TimescaleDB and Grafana, but we’d like to do it manually for now.
Two questions:
(1) How should we set up the schema for the database that will map well to what Grafana expects?
(2) How can we do that mappping from Timescale to Grafana? How does that “connection” take place? Is it just SQL queries pointed in the right direction?
In Grafana, you use the Postgres data source to connect to a Timescale database. Timescale maps pretty well to Grafana as it is time series data. It can be much harder to manipulate data from a standard Postgres table into the right format.
(2) How can we do that mappping from Timescale to Grafana? How does that “connection” take place? Is it just SQL queries pointed in the right direction?
The tutorial describes how to set up a connection using the standard Postgres “connector” (data source in Grafana terms): Timescale Documentation | Getting started with Grafana and Timescale
Once the data source is set up then it can be used in Grafana dashboards. A dashboard contains one or more visualizations (panels) and in those panels, you can choose your Timestream/Postgres data source as the source and then specify an SQL query.