Preliminary impressions of Salesforce Backup and Restore

In the last few weeks, I’ve had the opportunity to do some hands-on setup of the new Salesforce Backup and restore product. These are my initial impressions from that work.

Installation and setup

The installation path is standard. You install a managed package, assign some permissions and permission set licenses and off you go. We didn’t have any issues with the initial installation, it was quite smooth.

You access Backup and Restore via the App Launcher. The app contains tabs for backing up and restoring data, viewing logs, and viewing the configuration. The only real thing to configure is which AWS region you are going to store your backups in and ensuring that these endpoints are accessible from Salesforce.

This is the first important point: Your backups are held in AWS not on Salesforce, following an increasingly common pattern used by for instance Einstein Activity Capture. That means strictly speaking this isn’t native backup, although you will hear people refer to it as such.

Configuring a backup policy

The backup configuration is based off the objects you want to include and that is what you set up when you set up a backup policy. Using the interface, you pick which objects you want to make part of the policy and you select whether or not you want to include just the object itself or also related objects.

You save and activate the backup policy and it is put into a queue to be executed once every 24 hours. You have no control over the specific timing of the backup runs. In our tests, about 95% of objects backed up smoothly and the rest had errors of various kinds, some which are still under investigation. That is to say, this is a new product and has some teething issues that will no doubt be fixed in due time.

Restoring files

The restore interface is also quite simple. You start by selecting an object to restore and then the backup you want to restore from. You can only restore from specific backups, which can be a limitation if you don’t know when specific data was corrupted. ?

To check what will be restored, you can use the preview functionality to compare records. This gives you a field by field comparison of all fields that have changed since your chosen backup.

The restore functionality worked without issue in our tests, but the interface is not geared towards large scale bulk restore operations, as you need to mark files to be restored individually. It will take some time to mark up the files correctly, if you are restoring a large number of objects distributed among other files that aren’t meant to be restored.

Monitoring the backups

There are fairly detailed logs for both backup and restore operations that give and object by object status of the operations. You could wish for some more detailed debug logging at this stage of product maturity, but for most purposes it should prove adequate.

There are logs both for enqueue, start, and complete states, which means you can track the processes underway. That is handy as the time to complete a backup can be substantial.

Key takeaways

Overall, my impressions from these initial tests are as follows:

·??????Salesforce Backup and Restore is a new product with some quirks as a consequence. It is not a replacement for enterprise-grade solutions like Odaseva or Ownbackup, but it is a good choice if you’re looking for easy-to-configure daily backups

·??????For better or worse, the user interface and configuration are simple, with few options, and no bells and whistles

·??????Although, you will hear it referred to as native backup that is a truth with modifications as your data is actually stored to AWS

·??????Backup policies include the ability to include and exclude specific objects and their related objects. There is no control over backup schedules

·??????Restores work from specific backups and is geared more towards restoring individual files than bulk operations at this point in time.

·??????The built-in preview function allows you to compare the current state of records to those in your chosen backup

·??????There are fairly comprehensive logs that gives you good visibility into what is happening with the asynchronous operations. You could wish for more details on errors, but that will be less need as the product matures?