[gen220]

Is there an explainer for people who are broadly familiar with the DB space? It sounds like you're building an equivalent to Vitesse for Postgres, but it's not super clear from the article (which I know is not the point of this, but still :) ).

Edit: It also might be interesting to point out how your solution differs from what the folks at Planetscale are building https://planetscale.com/neki

[parthdesai]

There's multiple solutions coming up in this space:

1. Neki as you mentioned 2. PgDog 3. Multigres, headed by original creator of Vitesse

[frollogaston]

Citus is an older one that does something like this, right? But it's an extension, not a proxy.

[parthdesai]

I could be wrong, but with Citus, for most use cases, you can only have one co-ordinator node which fans out requests. So theoretically, you still can run into bottle necks at some point if 1 coordinator node is not enough.

With proxies like pgdog, multigres, and eventually Neki, these can scale out horizontally, so you get true unlimited scale.

[frollogaston]

Doesn't PgDog only have one proxy, or can you have multiple? I imagine they'd need to coordinate on sharding rules somehow.

[levkk]

You can have multiple. All sharding is config-based, so no real-time synchronization is required.