2 months ago
Hi all!
How does Railway ensure that automatic volume backups (i.e. for a MySQL server instance) have good data integrity? When you create backups from MySQL itself, it handles the synchronization to ensure that the data is in a consistent state, but as the automated backups happen on the volume itself, I am unsure if the at-rest data can be trusted (i.e. only capturing partial writes). Is there extra synchronization logic applied to the server instance itself, or is this something we need to manage ourselves?
Also, is it possible to configure the time at which these backups occur? We are evaluating Railway as a new platform for our services, moving away from dedicated hosts, and our users are aware of our current backup schedule and its associated slowdowns - if the backup process will impact our performance, we'd like to keep this the timing consistent if possible.
Thanks!
Thaum
3 Replies
2 months ago
Hello,
Backups are just ZFS snapshots, and in my years of doing support, I have never seen that cause any issues with in process writes or similar.
Backup times are not customizable, they are taken at the random time that is chosen once when the schedule was enabled.
Given backups are ZFS snapshots, there is absolutely zero performance degradation.
Best,
Brody
2 months ago
Oh it's ZFS, that's sweet - absolutely amazing. Resolves all of my concerns here! . Thank you! <:heartowo:626464756156596244>
2 months ago
No problem!
Status changed to Solved brody • 2 months ago