[Private Beta] Data Tiering to Amazon S3 in Timescale Cloud

Timescale Cloud now includes a consumption-based, low-cost object store built on Amazon S3, currently accessible via private beta.

This novel functionality expands the boundaries of traditional cloud databases, leveraging the flexibility of Timescale Cloud’s cloud-native architecture so you can store high volumes of time-series data for cheap. More specifically, by developing this feature, we aim to bring you:

  • :chart_with_upwards_trend: Bottomless scalability. Having native access to a cloud-native object store means that you can cost-efficiently store a bottomless amount of time-series data in Timescale Cloud. You can now create a data tiering policy and seamlessly tier data to this low-cost object store. Our goal is to save you the manual work of building and integrating up a custom system for your warehousing needs, or even operating a separate data store like Snowflake for archiving your time series.

  • :moneybag:Cost effective storage. This object storage layer is economical (around 10x cheaper than our regular storage) and consumption-based. You will pay only for what you store, with no extra charge per queries.

  • :computer:Query S3 with SQL directly from Timescale Cloud. After tiering data, hypertables will continue to look like a single, unified whole. By running one SQL command, you can move chunks to the low-cost object store; afterwards, you can still query your tables transparently using standard SQL, including JOINing this table alongside the rest of your tables. This avoids data silos + the extra operational complexity of having your data divided across multiple tools.

To learn more about this exciting new capability, read this blog post.

Request access

This feature is accessible to you as a private beta.

:warning: This means that this functionality is still under active development and not ready for production use. But as always, we’d like to rely on as much user feedback as possible before we launch this feature as generally available—we’d love for you to test it and give us feedback.

To become a beta tester,

  1. Sign up to Timescale Cloud if you haven’t already. The first 30 days are free, no credit card required.
  2. Log in to the Timescale Cloud UI. In your Service screen, navigate to Operations > Data Tiering. Click on the “Request Access” button.
  3. We’ll be in contact with you shortly to grant you access.

If you have any questions, please reach out to us! :pray:

Hi Carlota

This post specifically mentions Timescale Cloud in regards to Data Tiering.

The official docs Timescale Documentation | Data tiering doesn’t mention Cloud or Managed Services (which this forum is calling hosted services???).

So i tried the commands listed in the doc link, e.g. SELECT * FROM timescaledb_osm.tiered_chunks;

And I get ERROR: relation “timescaledb_osm.tiered_chunks” does not exist.

I don’t see any instructions on installing extensions? Do i need to sign up for something? Or is this feature only relevant for Cloud?

If that’s the case, and as a customer I’m really looking for both:

  • connection pooling (because we host in Lambda, and without it we’re totally screwed),
  • data tiering (because we are nervous about applying data retention policies in combination with materialized views which may need to be rebuilt)

what are our options?

Guess i’m going to answer my own question.

Just signed up for a Cloud account and data tiering is present in the UX.

So… guess i’ll contact support for an ETA on connection pooling in the Cloud offering - I was previously informed Q2 2023, so maybe it isn’t too far away which would be great as the pricing looks significantly better than Managed Services as well.