14.5K ·GitHub
What is this?
Litestream is a standalone disaster recovery tool for SQLite: a background process that safely replicates database changes incrementally so you can recover after a failure.
Litestream is a standalone disaster recovery tool for SQLite: a background process that safely replicates database changes incrementally so you can recover after a failure.So it's a one-off backup script, or something I keep running?You keep it running alongside your app — it streams changes incrementally as they happen, rather than taking snapshots on a schedule.
You keep it running alongside your app — it streams changes incrementally as they happen, rather than taking snapshots on a schedule.How do I get it running?The README sends you to the Litestream web site (litestream.io) for installation instructions and docs. The repo also ships a Dockerfile if you prefer containers.
The README sends you to the Litestream web site (litestream.io) for installation instructions and docs. The repo also ships a Dockerfile if you prefer containers.Does it modify my database in any way?Yes — it creates an internal litestreamlock table in the source database to coordinate WAL checkpoints. That changes the schema and page count, so the database is no longer byte-identical to its pre-Litestream state.
Yes — it creates an internal litestreamlock table in the source database to coordinate WAL checkpoints. That changes the schema and page count, so the database is no longer byte-identical to its pre-Litestream state.Can I run it in Docker or on Kubernetes?Both: the repo includes a Dockerfile, and community docs cover running Litestream on Kubernetes. It's a single Go binary, with OpenBSD release builds available too.
Both: the repo includes a Dockerfile, and community docs cover running Litestream on Kubernetes. It's a single Go binary, with OpenBSD release builds available too.Where do the replicas go — only S3, or other places too?The README names two destination types: another file or S3. Replication is incremental, so changes flow over continuously instead of as periodic full dumps.
The README names two destination types: another file or S3. Replication is incremental, so changes flow over continuously instead of as periodic full dumps.Could Litestream corrupt my database?The project makes that a stated design point: it only communicates with SQLite through the SQLite API, and per the README it will not corrupt your database.
The project makes that a stated design point: it only communicates with SQLite through the SQLite API, and per the README it will not corrupt your database.What happens if I drop the litestreamlock table while it's running?Checkpoint synchronization can fail until Litestream reinitializes the database. Restarting Litestream recreates the table automatically, but you shouldn't drop it mid-run.
Checkpoint synchronization can fail until Litestream reinitializes the database. Restarting Litestream recreates the table automatically, but you shouldn't drop it mid-run.If I restore a database, will the lock table end up in it too?Yes — because the source schema is backed up, restored databases also contain the table. It's expected and safe to leave in place.
Yes — because the source schema is backed up, restored databases also contain the table. It's expected and safe to leave in place.What are similar projects?velero, seaweedfs, kopia, sftpgo, artifact-server, chaosmonkey, ceph, juicefs
velero, seaweedfs, kopia, sftpgo, artifact-server, chaosmonkey, ceph, juicefsWhich topics is it in?replication, s3, storage, backup, aws, kubernetes, sqlite, cloud, database, go, tool, apache-2.0
replication, s3, storage, backup, aws, kubernetes, sqlite, cloud, database, go, tool, apache-2.0