Show HN: Postgres Platform for Global Replication and Sharding

Hi HN! This is David and Will from Fortress. Last month, we launched Fortress as a Postgres platform targeted at multi-tenant SaaS companies (https://news.ycombinator.com/item?id=41426998). Since then, we’ve pivoted our messaging to focus more on ‘globally distributed’ Postgres for developers.

Over the past month, we learned that for us, the multi-tenant messaging didn’t resonate as much as we had hoped with early users. However, users wanted a Postgres platform that made global distribution (replication and sharding) easy and cost-effective. This was particularly challenging for developers wanting to use global edge runtimes but were constrained by a single, centralized database.

Over the past few weeks, we’ve revamped the platform and shifted our focus, prioritizing the simplification of global replication (single-writer, multi-reader) and sharding for Postgres developers.

Here are some ways Fortress can be used, especially as the platform becomes more developed: 1. For developers (including API providers, e-commerce, and web developers) deploying their applications to the network edge to reduce latency for end users. 2. For developers who need to shard data in specific regions for compliance purposes. 3. For developers looking for a platform that provides the peace of mind of managed horizontal scalability and sharding.

Currently, the platform uses RDS under the hood and a proxy-based clustering system that supports write-forwarding and DNS routing that automatically directs queries to the closest replica. We haven’t made sharding self-serve yet, but we’re happy to assist with this over a call.

Our philosophy is to build on top of native Postgres without making changes to the database itself. We understand that naturally, building a middleware layer (shared-nothing) will naturally reduce Postgres compatibility, but we believe that it is important to stay close to native-Postgres to the extent that we can.

We’d love to hear feedback and suggestions from the HN community. We haven’t run a ton of benchmarks yet as we believe there is a lot we can do in the short-term to improve latency and we plan to do a Jepsen test as we continue working on the sharding solution. We are opening up our free tier, which includes a free primary and two replicas for super early users. You can use this from our site. If you’re interested in building on our platform, we’d love to chat and offer more granular support for sharding and help build out requested features (e.g., branching).