- Newest
- Most votes
- Most comments
-
- You can schedule backups as often as you like. Cron expressions can have day of week - see https://docs.aws.amazon.com/AmazonCloudWatch/latest/events/ScheduledEvents.html
-
- EBS looks after this. When a snapshot is deleted, data is first copied from it so that the next newer snapshot, despite being incremental, has everything it needs.
-
- Cold storage is available only for EFS, so for EC2/EBS there's no lifecycle out to cold storage. When a backup expires, it's deleted.
-
- You could have a look at your existing EBS backups to get storage estimates.
-
- "Start within", along with the start time, defines the window of time where the backup may start. We typically make this 1 hour to fit around other activities like maintenance windows. It's not such a big deal with EC2, but it can be tight fitting everything in with RDS as you often need RDS to be doing its own automatic backups as well (e.g. if you have a read replica) and your "AWS Backup" backup won't start if it's less than 4 hours before the automated backup window!
EBS snapshots aren't like old-school backups - other than your currently-oldest snapshot of a volume, they give you the storage benefits of being incremental while appearing to you in every way that they are full backups. So think of them all as full backups except when you look at your bill and see you're being charged for a lot less storage than you'd expect.
People have complained about how the console just shows the volume size as "Size" for every snapshot, not the true snapshot storage size. But maybe it's partly to encourage people to think of every snapshot as "full" because in every way you interact with them, they are.
Still, there is a way to get more info since the "EBS Direct" APIs were released in 2019. You can now get the amount of data blocks that changed from one snapshot to another. There's a basic script at https://github.com/akirsman/Snapshots/blob/master/snapshotsSize.py which I've used and enhanced. For example it reported for one of our volumes that the first snapshot is 50GiB (volume size) and the second is 0.859GiB.
For storage estimation you can also have a look in the Cost Explorer to see what your current snapshots are using.
Hello.
Thank you for your answer.
I understand the first backup is full, then every backup will be incremental for ever.?
When you say " When a snapshot is deleted, data is first copied from it so that the next newer snapshot, despite being incremental, has everything it needs", it means the first full backup will be deleted, but before, the data is copied to the next newer snapshot, so I will always have a full backup in my backup chain?. Per example if my retention is 30 days, every month I will have something like a full backup?
If I configure my backup rule for hourly, the options "start within" and "complete within" will be used?
Thank you.
Relevant content
- Accepted Answerasked 2 years ago
- asked 2 years ago
asked 3 years ago- asked 2 years ago
AWS OFFICIALUpdated a year ago- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
Hello.
Thanks for your answers.
1.- I understand that first backup will be ful, but all following backups will be incrementals, and there is not way to configure backup full every week, is that correct?
2.-I have configured a backup plan and rule for EC2 (disk1: 40GB, disk2: 20GB) with 02 days of retention, but I see that every the backup size is 60GB, so I understand that incremental backups do not work. How can I validate if incremental is working, or this is normal behavior, and I should see the size in another way?
3.-About storage estimation, I do not have EBS backups, maybe I have to consider a % of change, but the incremental backup should work, or the backup size for every day will be the sum of disks?