Looker Connect Training

Help Center

Documentation

Community

Cloud Certifications

Question

Connecting Looker to TimescaleDB

  • 30 November 2021
  • 2 replies
  • 68 views
M

Hey there,

I was wondering if anyone has experience with connecting a TimescaleDB (https://www.timescale.com) to Looker and can help me out with an experience report or any resources on that matter. Some questions I have:

  • Is a connection possible at all? I know that its not in the sql dialects listed by looker, but its basically a postgres database optimized for time series
  • If a connection is possible, are there any limitations to what queries you can run against it?
  • Can you use timescale specific functions like time buckets when defining views?

I will take it for a test drive anyway, but I though maybe someone already did that before me and is willing to share some insights. If not, I’m happy to share my learnings from it.

Best, Markus
This topic has been closed for comments

2 replies

greenweeds

Hello

Maybe someone else in the community has direct experience but… 

You can connect a PostgreSQL database to Looker which means TimescaleDB should work properly as well. Timescale is an extension to PostgreSQL and you should be able to take a standard PostgreSQL approach to connnections etc. 

As for the TimescaleDB specific features: e.g continuous aggregates should work as a derived table (based on Looker's docs). You should be able to use TimescaleDB hyperfunctions like time_bucket, as Looker supports custom SQL queries (Looker docs). In some cases, you might need to write your own custom SQL, and not use Looker's no-code tool. 

Transparency: I’m Timescale’s community manager, this is based on assumptions about the way that Looker works with PostgreSQL. Would be happy to hear how you get on. 

M

Thanks for the reply. I also believe that it should work. Here is the official reply from Looker Support:

 

I asked about the TimescaleDB and I have received confirmation that connecting and using it should be possible as it is PostgresSQL based.

To clarify one thing, we don't officially support TimescaleDB however so if anything is breaking with it that isn't PostgresSQL then there isn't much we can do to fix it. I would recommend putting in a feature request on Pendo for TimescaleDB support so our Product team can see that it is in demand.

 

I’ll keep you posted ;)

Best, Markus
Terms & Conditions