CedarDB Database Upgrade

Stop concurrent access

To prevent data loss, make sure that no other process is writing data to CedarDB while the upgrade is performed. The data export runs in a separate transaction to ensure a consistent state of all your data, but any concurrent changes during the export will be lost!

Copy data to CSV files

Once you have stopped all writer processes, copy all data out of CedarDB into CSV files.

First, connect to CedarDB using the psql tool from PostgreSQL.

psql -h localhost -U {{username}}

Following, copy all your tables to CSV files by running the following command for each table:

\copy {tablename} TO 'your/backup/path/{tablename}.csv' DELIMITER '|' CSV NULL '';

Stop all running CedarDB instances

First, stop all running CedarDB instances.

Backup and move old database files

To ensure access to the old database in case of problems during the upgrade, create a backup of the database directory. This can be removed once the new version is running and all data has been imported correctly. Once all running CedarDB instances are stopped, this can be done safely by any method, e.g., using the mv command.

mv your/path your/path.bak

Install latest version

Please follow the installation guide to install and start the latest CedarDB version.

Re-create your schema

Connect to the CedarDB instance and re-create your schema, e.g., by running your schema.sql script:

psql -h localhost -U {{username}} < schema.sql

Copy data from CSV files

Once your schema is created, copy the contents of all tables from the CSV backups you created earlier. Run for each table:

\copy {tablename} from 'your/backup/path/{tablename}.csv' DELIMITER '|' CSV NULL '';