3 ms·
Two things to learn from this: 1. Never have a single point of failure. Relying on DO for Server+Backups is putting your eggs on one basket. 2. Your server st
by csomar 11y ago
Two things to learn from this:
1. Never have a single point of failure. Relying on DO for Server+Backups is putting your eggs on one basket.
2. Your server state should be programmable. This is not quite easy for complex configurations. But today, DO has an API, we have Docker, and quite modern deployment tools.
Here is setup:
1. Github for the server state. Basically, a repository to configure and deploy my infrastructure.
2. Enable DO backups in case I mess up something and want a quick come-back.
3. File backups through Tarsnap. Since I use Docker, I have a volume container. Backup the volume container with Tarsnap.
- peterwwillis 11y agoThere's always a single point of failure. Most commonly it's a human.