Scenario: a colleague accidentally ran 'rm -rf' on the database master node directory. What should be done?

First, it is necessary to stop the database to prevent further data damage. Then:

1. Assess the situation: check which data and files were deleted.
2. Restore from backups: if current backups are available, restore the master node directory from them.
3. If backups are not available, try using data recovery tools, but the chances of successful recovery are low.
4. After recovery, check the integrity and functionality of the database.
5. Implement measures to prevent similar incidents in the future: restrict access rights, set up protection against accidental deletion, automate backups.

It is important to act quickly but carefully to avoid worsening the situation.