For some customers, SaaS is simply not an option. The ability to manage Materialize within private infrastructure - a self-managed, or self-hosted, solution - has been our prospects’ number one request. Many companies were excited by Materialize’s potential to use SQL to transform, deliver, and act on fast-changing data, but they needed:
- To meet governance, risk, and compliance requirements that made SaaS procurement difficult or impossible, and/or
- To run Materialize in cloud providers and regions not supported by our existing managed service.
So we listened – and we now have an early access version of self-managed Materialize, with the first production release slated for early 2025.
New look, same great compute
While our self-managed solution is new, the underlying technology remains the same: A well-honed, powerful operational data store (ODS) that creates and continually updates consistent views across your organization’s data.
In other words, the ingredients inside self-managed Materialize are the same great ingredients inside our managed cloud product: Crazy-fast compute clusters powered by differential dataflow; a strict serializable SQL interface; and a mature cloud operator.
Just as our managed cloud customers have been doing for years, you can now use Materialize to simplify your architecture within your private infrastructure:
Simplify your architecture with three common patterns enabled by Materialize | ||
|
Scale complex read queries more efficiently than a read replica, and without the headaches of cache invalidation. |
Extract, load, and incrementally transform data from multiple sources. Create live views of your data that can be queried directly or pushed downstream. |
Use SQL to create and deliver real-time, strongly consistent data products to streamline coordination across services and domains. |
What’s in the box?
We’ve designed self-managed Materialize for maximum flexibility, meaning that in all likelihood you will be able to simply download, plug it right into your existing infrastructure, and start using the product.
What we give you: We’ve packaged up Materialize into a Helm chart that can be installed in your Kubernetes cluster. The Helm chart installs a full-fledged operator that provisions all of the necessary compute resources to operate Materialize on your private infrastructure. It’s the same operator we use internally to run our own services, so you can trust it’s battle-tested and built for reliability at scale.
What you need: A Kubernetes cluster, a Postgres database, and some blob storage. (If you don’t already have them, each of these prerequisites is only a click away in AWS, GCP, or Azure.)
Once you’ve downloaded the Helm chart, plugged in your blob storage and Postgres, and installed it on your Kubernetes cluster, you can have Materialize up and running within minutes.
Now it’s time for Materialize to show you what it can do.
Challenge 1: Ingest data from all your disparate, previously-siloed data sources: Kafka, Postgres, MySQL, or webhooks.
Challenge 2: Write that SQL query you’ve always wanted to run that joins across all these data sources - the query that was downright impossible for any other database you’ve tried. Don’t hold back - use every JOIN you want!
Challenge 3: Now want a fast version of that previously impossible query? CREATE INDEX on the query. Once it’s built, try running your query again. And again. And again.
Challenge 4: Make a change to an upstream data source. See if you can rerun your query faster than Materialize can update its results (hint: you’re going to need to be really fast 😉)
Challenge 5: Try running SUBSCRIBE on that SQL query that you’ve always wanted to be up to date, but was too complex or too slow for other databases to keep up. Watch as every update to the result is produced in real time. Sit slowly back as you envision the possibilities.
Adaptability wins
We believe Materialize breaks ground in real-time data transformation and unlocks a new world of opportunities for live data services. We’ve seen it work for our existing cloud customers - in use cases from logistics to manufacturing to financial services - and our new self-managed offering now makes Materialize accessible to an even wider audience.
Don’t settle for operating on stale data, join our early access program today and let Materialize push fresh, transformed data directly to you, in any cloud of your choosing.
