July 21, 2025

Recovery boot of a DBaaS Instance after a root filesystem filecorruption

Log into console

Navigate to Oracle base database service
locate the relevant Instance.
Select Console Connection
Create Console Connetion-generate keys.
Copy the SSH String.

Open cloud Shell

Create the private key within cloud shell.
Paste the SSH String and Connect.
Reboot the dbaas Instance
From Cloud Shell you will see the server booting.
When the robot Starts you will see a boot manager menu. Highlight The Top boot option and press e
In edit mode, use the down arrow key to scroll down through the entries until you reach the line that starts with the following value, depending on the version of Linux:

For instances running Oracle Autonomous Linux 8.x, Oracle Autonomous Linux 7.x, Oracle Linux 8.x, and Oracle Linux 7.x, scroll down until you reach the line that starts with linuxefi.

For instances running Oracle Linux Cloud Developer 8.x, scroll down until you reach the line that starts with linux.

At the end of the line add the following: init=/bin/bash
reboot the instance by using the keyboard Shortcut CTRL+X wait for the Instance to boot into Bash
Enter /usr/sbin/load_policy -i
Enter the following: /bin/mount -o remount, rw /
Make changes to the configuration files as needed…
To complete the recovery enter exec /usr/lib/systemd/systemd

Conclusion

Bringing a DBaaS instance back from root filesystem corruption is straightforward once you know the steps. Console access, single-user mode, and a few basic file edits are usually all that’s needed to get the system running again without resorting to a full rebuild. It’s not something you’ll need often, but when your instance locks you out and standard access fails, this method offers a simple, reliable way to get control back without overcomplicating the recovery process.

[email protected]