Immutability Duration and Retention usage advice

[and_ita]

Hello,
I've configured a daily incremental backup Job that save backups on a SOBR.
SOBR is composed by an onprem NAS and a S3 locked bucket.
The backup Job is configured with a retention policy of 7 days, backups are copied to object storage as soon as they are created.
The bucket settings are the following:
- make recent backups immutable for 30 days
- use infrequent access storage
- store backups in single availability zone
Backups works without problems but there's one thing that I don't understand. When I try to recover a backup I expect to find 7 restore points if I restore from the NAS and 30 restore points if I restore from S3 bucket because it is immutable for 30 day. But only 6 restore points are present. So I try to put the NAS offline and to rescan the S3 bucket but the restore points found are always 6. If the bucket are locked why Veeam cannot see the previous restore points?
Second question, I don't know if you can help me with this but I see that after 7 days from the beginning of this job the expense from S3 are increasing with a new cost called "EUS1-Requests-Tier1", do you have an idea of what this cost could be?

Thanks in advance for your help

[Mildur]

Hi Andrea

Welcome in the forum.
In a SOBR, restore points are stored accordingly to the configured retention. With a retention of 7 days, you will have 7 days of backups.

When I try to recover a backup I expect to find 7 restore points if I restore from the NAS and 30 restore points if I restore from S3 bucket because it is immutable for 30 day.

The following settings are required to fulfill your expectations:

Job:
- Short-term Retention: 30 days
- GFS Retention: weekly full

Object Storage Immutability:
- 30 days

SOBR Settings:
- Capacity Tier Policy: Move and Copy
- Operational Restore Window: 7 days

With that configuration, you will have 7-14 daily restore points on your NAS, and 30-35 daily restore points on your S3 bucket.
The copy policy will copy all restore points immediately to the S3 bucket. The move policy will move inactive backup chains to the S3 bucket. The weekly full backup is required to be able to move backup files.

Alternative Option with Veeam Backup & Replication v12:
Instead of using the S3 bucket in a capacity tier, use a backup copy job to copy backups from the NAS directly to the S3 bucket. The backup copy job should be configured with 30 days retention.

Second question, I don't know if you can help me with this but I see that after 7 days from the beginning of this job the expense from S3 are increasing with a new cost called "EUS1-Requests-Tier1", do you have an idea of what this cost could be?

Amazon writes on their userguide the following description for that costs:
region-Requests-Tier1 - The number of PUT, COPY, or POST requests for STANDARD, RRS, and tags, plus LIST requests for all buckets and objects

Those API calls are necessary. For example new unique objects are copied to the capacity tier in offload sessions. Object lock date for those objects will be updated every few days.

Best,
Fabian

[and_ita]

Thanks for your reply.
I have Backup and Replication V11.
When you said:
"- Capacity Tier Policy: Move and Copy
- Operational Restore Window: 7 days"

Do you mean that in SOBR I have to check "copy immediately to object storage" (current checked) and also "move backup to object storage as they aged out.." and set to move files older than 7 days?

[Mildur]

Yes. You expect to find 30 restore points on S3 (capacity tier), but only 7 on your NAS (performance tier).
In a Scale Out Repository, the Move Policy is the only way to achieve your expectation.

Copy to have a immediate copy of your backups on S3, and Move Policy to remove all inactive backup chains older than 7 days from the NAS.

Best,
Fabian