Exploring Built-in SeekDB Sync / Replication Support Inspired by SQLite Sync

[fkuner]

Hi everyone,

I recently came across SQLite Sync from SQLite Cloud, and I think the concept is really compelling—especially for applications that rely on SQLite but need lightweight, real-time synchronization across devices or instances.

This got me thinking: would it make sense to add similar sync/replication capabilities directly into SeekDB?

The idea would be to enable SeekDB users to:

• Keep multiple SeekDB instances (e.g., on different clients or edge nodes) in sync automatically.
• Handle conflict resolution gracefully (e.g., via timestamps, CRDTs, or user-defined logic).
• Maintain SeekDB’s simplicity and performance while adding just enough infrastructure for reliable sync.

I’m curious to hear your thoughts:

• Is this a feature you’d find useful?
• What sync model would work best for SeekDB (e.g., client-server, peer-to-peer, eventual consistency)?
• Are there existing libraries or protocols we should consider integrating (like Litestream, Dqlite, or a custom solution)?

This could open up new use cases for SeekDB in offline-first apps, distributed edge computing, or collaborative tools. Let’s discuss!

[PsiACE]

I'm not very familiar with seekdb yet, but there have been some interesting solutions in the SQLite community in the past:

• https://github.com/benbjohnson/litestream
• https://github.com/vlcn-io/cr-sqlite

[fkuner]

That’s super helpful—thanks a lot! 🙌

[fkuner]

https://github.com/powersync-ja/powersync-sqlite-core

This project seems to be doing the same thing.