I've been thinking about whether we really need an open source alternative to tools like Kasten that can effectively recover your CNPG cluster and also automate disaster recovery drills. I previously asked a similar question in another community and didn't get much feedback. My vision is to have something that sends me a report confirming that our organization is ready to handle any disaster. While I've seen many projects focused on backups, I want to highlight the importance of guaranteed restores and automated DR exercises as well.
3 Answers
Absolutely! An open source solution could be really beneficial, but it needs to seamlessly integrate with existing operators. For instance, while I can manually recover a database, having a user-friendly UI would make it a lot easier to manage that process.
You're right! Automation is key, but managing all these scripts can become overwhelming. Tools should ideally help simplify things, especially when it comes to the know-how required for upgrades and potential pitfalls.
You could definitely set up a job to create a new CNPG cluster from a backup and run some tests to confirm that the recovery works. It’s a practical solution, but I feel like I've got too many scripts for different tasks already. It raises the question: if we can script everything, do we really need these tools? Plus, with upgrades, there are plenty of challenges to navigate.
Totally agree! It should leverage the capabilities we already have. I think having yet another operator is unnecessary. However, it should address critical features, like ensuring that backups are also backed up to protect against theft or corruption.