Skip to main content

Why Can Cloud Incremental Backups be deleted, without effecting the backup chain?

Thread needs solution

It's my understanding that cloud backups use an incremental strategy. Per the incremental strategy guide:

https://www.acronis.com/en-us/support/documentation/ATI2023/index.html#…

"Therefore, if you lose an incremental backup version or it becomes corrupted, all later incremental versions are unusable."

 

Why is this not true for cloud incremental backups? Is there some processing done, to revise the full backup with the deleted incremental backup's data?

0 Users found this helpful

Jacob, welcome to these public User Forums.

There are fundamental differences between local incremental backups and those stored in the Acronis Cloud.

Local backups use a version chain scheme, i.e. an initial Full backup followed by a number of Incremental 'slices' (as Acronis refers to them) which with TIBX format files are automatically consolidated in to the same file as the initial Full backup.  This makes the chances of losing an individual incremental slice almost impossible without the whole TIBX file having been corrupted in some way.  This scheme is normally repeated multiple times as new version chains are created and old ones deleted.

Cloud backups only ever create an initial Full backup which is uploaded to the cloud servers, thereafter they use a hybrid delta / incremental process that tracks only changes to the Source data for the task.

In this way, you are actually using a form of versioning, where you can select the criteria in the Advanced Options for the task for how many versions over what period of time are retained before the oldest version of that data is removed in favour of keeping a later version.

This versioning process for cloud backups will work fine but may result in some significant growth in version sizes if the backup is of the Windows OS disk and you subsequently upgrade the OS version, i.e. from Windows 10 to 11 where a very large volume of files in the source data are replaced.  That should normally cause the user to consider creating a new backup task for the new OS version data, assuming that they have sufficient available storage capacity for doing so in the cloud.

Thanks for posting a detailed answer. It sounds like cloud backups are really differential backups. Where new backups are always compared to the initial full backup.

Would you say that's a correct statment?

Jacob, they are not differential or incremental in terms of how these work when used with local backup destinations.

Differential backups tend to grow in size with each new backup slice as this will always contain all changes since the initial full backup was created.

The key intention with cloud backups is to try to minimise the amount of data that needs to be uploaded once the initial full upload has been completed, hence the delta type approach and use of versioning to try to optimise the use of cloud storage while maintaining the integrity of the data being protected.