Go to file

Ben Johnson d4891f33da Refactor DB.checkpoint() to accept any mode.

This was originally meant to add a TRUNCATE checkpoint before starting
a new generation, however, there is a write lock that blocks the
checkpoint and it's more complicated to roll it back and attempt the
truncation.

2020-12-29 17:02:33 -07:00

.github/workflows

Add release workflow

2020-12-29 13:44:58 -07:00

cmd/litestream

Add release workflow

2020-12-29 13:44:58 -07:00

doc

Remove fuse

2020-12-17 15:15:01 -07:00

db.go

Refactor DB.checkpoint() to accept any mode.

2020-12-29 17:02:33 -07:00

go.mod

Change config format to yaml; add replicators

2020-12-18 13:21:29 -07:00

go.sum

Change config format to yaml; add replicators

2020-12-18 13:21:29 -07:00

LICENSE

GPL v3

2020-11-06 08:40:50 -07:00

litestream_test.go

Fix unit tests.

2020-12-29 12:56:50 -07:00

litestream.go

Refactor DB.checkpoint() to accept any mode.

2020-12-29 17:02:33 -07:00

README.md

Add read lock

2020-12-19 08:56:44 -07:00

replicator.go

Group WAL files in file replica

2020-12-29 16:40:28 -07:00

README.md

litestream

Streaming replication for SQLite.

Questions

How to avoid WAL checkpointing on close?

Notes

-- Disable autocheckpointing.
PRAGMA wal_autocheckpoint = 0

Languages

Go 98.8%

Makefile 0.5%

Python 0.3%

Dockerfile 0.2%

PowerShell 0.1%