Back to topics

EU Postgres in Focus: Choosing Managed HA Clusters Amid Vendor Concerns and New Tooling

1 min read
229 words
Database Debates Postgres Focus:

Europe-based managed HA PostgreSQL is at a crossroads. An Austria-based deployment needs a resilient 3-node cluster with 1 TB up to 6 TB total, and per-node specs around 8 vCPU and 32–64 GB RAM. Geopolitical concerns have teams eyeing EU hosts, with OVH highlighted as a strong option [1].

EU Hosting Landscape EU hosting options for this setup include OVH, seen as a credible European option for a 3-node HA Postgres cluster. The plan centers on 1 TB → 6 TB capacity and per-node resources around 8 vCPU and 32–64 GB RAM, with the Austria base shaping latency considerations [1].

Vendor Considerations The thread flags a push away from the “big three” due to geopolitical risks, nudging teams toward EU-native choices [1].

Edge Tooling Trends Two Bun-native packages are making Postgres glue code obsolete:

  • trpc-bun — Bun-native tRPC adapter: HTTP + WS. Runs on Bun.serve and uses a fetch adapter plus server.upgrade; includes duplicate-id protection and graceful disconnect. First discussed as a Bun-native tRPC bridge [2].

  • kysely-bun-sql — Kysely Postgres dialect backed by Bun SQL; provides pooled SQL under the hood, full Kysely integration (adapter, query compiler, introspector), transactions and savepoints, with no runtime deps and ES Module-only design [2].

As EU hosting and edge tooling converge, European Postgres usage looks increasingly edge-friendly, with OVH as a viable EU anchor and Bun-powered tooling accelerating near-edge deployments [1][2].

References

[1]
HackerNews

Ask HN: Where do you host managed HA PostgreSQL clusters in the EU?

Discusses options for managed HA PostgreSQL in Europe, provider selection, concerns about large vendors, considers OVH; Austria specifics.

View source
[2]
HackerNews

Two tiny Bun-native packages tRPC over Bun.serve and a Kysely Postgres dialect

Announcement of tRPC Bun adapter and Kysely Bun SQL dialect for Postgres; aims ease, no Node shims, edge cases welcome.

View source

Want to track your own topics?

Create custom trackers and get AI-powered insights from social discussions

Get Started