How to: Off-site Backup Job Best Practice Guide

This article will run you through Probax's recommendations when backing up to the Probax VCC Cloud environment

1. Job Naming Conventions
2. Retention
3. GFS and Archival
4. Backup Health Checks
5. Periodic Full Backups
6. Scheduling
7. Encryption
8. Configuration backups

1. Job Naming Conventions

Veeam's Cloud Connect functionality works on complex data basing that requires a few pre-requisites to be followed. We recommend keeping job names simple and instructive, allowing for engineers at a glance to understand what is being backed up and where the backups are destined to go.

• Best practice for off-site job names is to avoid special characters and utilize underscores to separate words in the job's name (e.g “bobsprinting_probax_offsite”. 
• For customers using a backup job to send over a local job's data, please maintain the same nomenclature and style:
  
  • Local job name example: "bobsprintingDC_local_job"
  • Cloud job name example: "bobsprintingDC_offsite_job"
• Please do not use spaces in your job's name, opting for underscores instead.

2. Retention Settings

Retention in Veeam ultimately determines your RTO and storage usage, so assigning the correct retention is critical. 

• Partners copying local backup data to an offsite location need to decide between 'days' and 'restore points':
  • Setting retention using 'days' tells Veeam to keep as many day's worth of images that you set. This means, if your local backup job is running 6 times a day for a 7 day retention, your cloud retention is 42 restore points
  • Setting retention using 'restore points' tells Veeam to only keep that many files on the cloud. Customers backing up multiple times per-day will need to re-think using this option
• Backup jobs can give you the option to chose between:
  • "Mirroring” - This is recommended for customers wanting to push up the same data locally, as they desire on the cloud. Probax recommends this for customers with larger, more powerful backup environments running per-day, with little limitation/throttling on networking needed
  • “Pruning” - This causes Veeam to send a single file per-day. Probax recommends using this for customer's who's backup window may be narrow, or who want to keep a different retention on the cloud

3. GFS and Archival

For partners looking for long-term, air-gapped storage, utilising archival retention may be the correct solution, instead of having an extensive retention period set (on regular hot storage backups). Please review the following KnowledgeBase area to learn more about this solution. 

4. Backup Health Checks

Probax recommends leaving health checks off for off-site backup jobs, due to the increased load these present to the backup server. Whilst health checks are a positive way to ensure your backups are healthy, the trade-off involving increased iops and network usage (and therefore, increase chance of interuption to regular backup schedule) doesn't balance out the benefits ensued.

5. Periodic full backups

As with health checks, Probax recommends off-site jobs do not have regular compact/active/synthetic full backups enabled. Customers can see interruption to the backup schedule if these are enabled.

6. Backup Schedule

Probax recommends weighing up the network/iops load for the customer with the frequency/timing of backups. It's important to consider when local backup jobs are running aswell, as despite the main consideration for cloud jobs being network-based, local jobs can cause delays too. 

7. Encryption

Though it is recommended to encrypt local backup job data, Probax recommends considering leaving encryption off for off-site jobs. This is because traffic between the backup site and the VCC repository is encrypted by Veeam regardless, where setting encryption in the job itself simply applies authentication to the restore points.

8. Configuration Backups

Probax recommends ensuring these have been enabled in backup jobs utilising Veeam Backup and Replication. These will assist in cases of manual restoration, where a Veeam server will need to be backed up from scratch. Veeam have more information in their kb.