Replies: 1 comment 3 replies
-
|
If you run Litestream as a sidecar (as this seems to be compose) you must ensure the application SQLite doesn't roll in the WAL file which happens automatically when the last connection to the database is closed. To prevent that from happening in sidecar mode your application needs to open the database with SQLITE_FCNTL_PERSIST_WAL option. That might be hard to do as it's a very low level setting. The easier approach is to use |
Beta Was this translation helpful? Give feedback.
-
|
Thank for your answer, I will try. In the meantime, I tried to backup on filesystem ( no type in config, only path ). And I didn't get any error and the restore process was correct. Shouldn't I have had the same in this case ? |
Beta Was this translation helpful? Give feedback.
-
|
It's replication that's affected by it, restore downloads a fresh copy every time anyway. |
Beta Was this translation helpful? Give feedback.
-
|
Sorry, I'm having trouble making myself understood. I made a test with this following config (no S3 type) : I didn't have the error "monitor error: cannot determine replica position: no snapshot available ..." when I use replicate command. Shouldn't I have had the same in this case ? Moreover, I notice that in this "filesystem mode", when I make a ls on /home/test/audit-db/generations/XXXXXX I have a "snapshots" directory AND a "wal" directory. In S3 mode, I have only on subdir "snapshots" in "generations/XXXXX/" |
Beta Was this translation helpful? Give feedback.
-
I try to backup a database on a S3-compatible storage. When I start a docker container with litestream, I obtain this error:
On object storage, I find some backup files like
What "cannot determinate replica position error" message means and how can I solve it ?
Thanks for your help
Beta Was this translation helpful? Give feedback.
All reactions