I know I am late on the trend, but due to planned upgrade to Veeam B&R 13 we need to change our backups.
My current backups are reverse incremental, with 35 restore points and active full the first monday of the month.
We check for backup integrity, but do not use SureBackup. Backup is to DAS and we do have the repository on NTFS and uses 15 TB (small fish, I know).
We have 14 jobs, because we have different backup windows.
I planned to move to 35 days and incremental backup with monthly active full, but I am worried that we will have time when we have almost 2 months of restore points for each backup.
Can you recommend a backup configuration that will stay close to what we have with our reverse incremental backup?
I was burned by incremental backups in Veeam, but I admit it was several years ago, so I am a bit afraid of the change.
Suggestions and point of attemption based on your experience are more than welcome.
Thanks in advance
-
npitacco
- Influencer
- Posts: 19
- Liked: never
- Joined: Jan 12, 2010 3:33 pm
- Full Name: Nadia Pitacco
- Contact:
-
vnikiforov
- Veeam Software
- Posts: 213
- Liked: 68 times
- Joined: Aug 17, 2022 5:03 am
- Full Name: Vladimir Nikiforov
- Location: Romania
- Contact:
Re: Moving from reverse inclemental to Incremental
Hello, Nadia,
With forward incremental and monthly fulls, retention deletes a chain only as a whole, so there will be time when you would keep two months of restore points on disk. This is expected behavior, see Forward Incremental Backup Retention Policy.
The configuration closest to your current one is forever forward incremental: same schedule, 35 days retention, no periodic fulls at all. Once retention is reached, the job merges the oldest increment into the full at the end of each run, so the footprint stays at one full plus 35 increments - essentially what your reverse incremental gives you today. See Forever Forward Incremental Backup Retention Policy for how the merge works. It also runs after the VM snapshot is released, so the system load overhead is lower than when using reverse incremental.
The switch will not double your space usage: the existing full backup of the reverse incremental chain becomes the starting point of the new chain, and retention removes the old rollback files as usual - see Switching Between Backup Methods.
If you decide to go without periodic fulls you lose the integrity check your monthly active full, so I'd recommend to cover it with the job maintenance settings instead: schedule a monthly health check and enable Defragment and compact full backup file (the compact temporarily needs free space of about one full backup).
If you would rather keep monthly fulls, the temporary second chain on NTFS cannot be avoided - only a Fast Clone capable file system such as ReFS makes periodic synthetic fulls cheap in space. Something to consider. Here is relevant thread which might help you
- similar question but for ReFS, it answers how synthetic fulls + fast clone differs from NTFS case.
With forward incremental and monthly fulls, retention deletes a chain only as a whole, so there will be time when you would keep two months of restore points on disk. This is expected behavior, see Forward Incremental Backup Retention Policy.
The configuration closest to your current one is forever forward incremental: same schedule, 35 days retention, no periodic fulls at all. Once retention is reached, the job merges the oldest increment into the full at the end of each run, so the footprint stays at one full plus 35 increments - essentially what your reverse incremental gives you today. See Forever Forward Incremental Backup Retention Policy for how the merge works. It also runs after the VM snapshot is released, so the system load overhead is lower than when using reverse incremental.
The switch will not double your space usage: the existing full backup of the reverse incremental chain becomes the starting point of the new chain, and retention removes the old rollback files as usual - see Switching Between Backup Methods.
If you decide to go without periodic fulls you lose the integrity check your monthly active full, so I'd recommend to cover it with the job maintenance settings instead: schedule a monthly health check and enable Defragment and compact full backup file (the compact temporarily needs free space of about one full backup).
If you would rather keep monthly fulls, the temporary second chain on NTFS cannot be avoided - only a Fast Clone capable file system such as ReFS makes periodic synthetic fulls cheap in space. Something to consider. Here is relevant thread which might help you
- similar question but for ReFS, it answers how synthetic fulls + fast clone differs from NTFS case.
---
BR,
Vladimir
Veeam Software
BR,
Vladimir
Veeam Software
Who is online
Users browsing this forum: Amazon [Bot], Semrush [Bot] and 40 guests