bin/cli/docs/copy-database.md
Copy the Komodo database contents between running, mongo-compatible databases. Can be used to move between MongoDB / FerretDB, or upgrade from FerretDB v1 to v2.
services:
copy_database:
image: ghcr.io/moghtech/komodo-cli
command: km database copy -y
environment:
KOMODO_DATABASE_URI: mongodb://${KOMODO_DB_USERNAME}:${KOMODO_DB_PASSWORD}@source:27017
KOMODO_DATABASE_DB_NAME: ${KOMODO_DATABASE_DB_NAME:-komodo}
KOMODO_CLI_DATABASE_TARGET_URI: mongodb://${KOMODO_DB_USERNAME}:${KOMODO_DB_PASSWORD}@target:27017
KOMODO_CLI_DATABASE_TARGET_DB_NAME: ${KOMODO_DATABASE_DB_NAME:-komodo}
Up to Komodo 1.17.5, users who wanted to use Postgres / Sqlite were instructed to deploy FerretDB v1. Now that v2 is out however, v1 will go largely unsupported. Users are recommended to migrate to v2 for the best performance and ongoing support / updates, however the internal data structures have changed and this cannot be done in-place.
Also note that FerretDB v2 no longer supports Sqlite, and only supports a customized Postgres distribution. Nonetheless, it remains a solid option for hosts which do not support mongo.
Also note, the same basic process outlined below can also be used to move between MongoDB and FerretDB, just replace FerretDB v2 with the database you wish to move to.
Don't forget to also add the new volumes.
## In Komodo compose.yaml
services:
postgres2:
# Recommended: Pin to a specific version
# https://github.com/FerretDB/documentdb/pkgs/container/postgres-documentdb
image: ghcr.io/ferretdb/postgres-documentdb
labels:
komodo.skip: # Prevent Komodo from stopping with StopAllContainers
restart: unless-stopped
# ports:
# - 5432:5432
volumes:
- postgres-data:/var/lib/postgresql/data
environment:
POSTGRES_USER: ${KOMODO_DB_USERNAME}
POSTGRES_PASSWORD: ${KOMODO_DB_PASSWORD}
POSTGRES_DB: postgres # Do not change
ferretdb2:
# Recommended: Pin to a specific version
# https://github.com/FerretDB/FerretDB/pkgs/container/ferretdb
image: ghcr.io/ferretdb/ferretdb
labels:
komodo.skip: # Prevent Komodo from stopping with StopAllContainers
restart: unless-stopped
depends_on:
- postgres2
# ports:
# - 27017:27017
volumes:
- ferretdb-state:/state
environment:
FERRETDB_POSTGRESQL_URL: postgres://${KOMODO_DB_USERNAME}:${KOMODO_DB_PASSWORD}@postgres2:5432/postgres
...(unchanged)
volumes:
...(unchanged)
postgres-data:
ferretdb-state:
The SOURCE_URI points to the existing database, ie the old FerretDB v1, and it depends
on whether it was deployed using Postgres or Sqlite. The example below uses the Postgres one,
but if you use Sqlite it should just be something like mongodb://ferretdb:27017.
## In Komodo compose.yaml
services:
...(new database)
copy_database:
image: ghcr.io/moghtech/komodo-cli
command: km database copy -y
environment:
KOMODO_DATABASE_URI: mongodb://${KOMODO_DB_USERNAME}:${KOMODO_DB_PASSWORD}@ferretdb:27017/${KOMODO_DATABASE_DB_NAME:-komodo}?authMechanism=PLAIN
KOMODO_DATABASE_DB_NAME: ${KOMODO_DATABASE_DB_NAME:-komodo}
KOMODO_CLI_DATABASE_TARGET_URI: mongodb://${KOMODO_DB_USERNAME}:${KOMODO_DB_PASSWORD}@ferretdb2:27017
KOMODO_CLI_DATABASE_TARGET_DB_NAME: ${KOMODO_DATABASE_DB_NAME:-komodo}
...(unchanged)
Run docker compose -p komodo --env-file compose.env -f xxxxx.compose.yaml up -d, filling in the name of your compose.yaml.
This will start up both the old and new database, and copy the data to the new one.
Wait a few moments for the copy_database service to finish. When it exits,
confirm the logs show the data was moved successfully, and move on to the next step.
In your Komodo compose.yaml, first comment out the copy_database service and old ferretdb v1 service/s.
Then update the core service environment to point to ferretdb2.
services:
...
core:
...(unchanged)
environment:
KOMODO_DATABASE_ADDRESS: ferretdb2:27017
KOMODO_DATABASE_USERNAME: ${KOMODO_DB_USERNAME}
KOMODO_DATABASE_PASSWORD: ${KOMODO_DB_PASSWORD}
Repeat the same docker compose command as before to apply the changes, and then try navigating to your Komodo web page.
If it works, congrats, you are done. You can clean up the compose file if you would like, removing the old volumes etc.
If it does not work, check the logs for any obvious issues, and if necessary you can undo the previous steps to go back to using the previous database.