This article details what is 'Recommended' and 'Not recommended' within a Backup & Restore Job within Gearset.
'Recommended objects' - included by default in your backup
By default, each backup run will contain your organization's metadata, plus the data contained within:
Your custom objects (including any new objects created over time).
Managed packages (including configuration for CPQ, Vlocity, Conga etc.)
This will cover the majority of objects we can retrieve from your Salesforce organization. However, in a few cases there will be objects we do not recommend but which might be business-critical for your use-case.
'Not recommended objects' - excluded by default from your backup
Our recommended filter excludes the following sets of objects:
Associated objects (including Feed, History, Share etc.)
'Debug' objects (including Log, Metric, Usage etc.)
A full list of not recommended objects for your organization can be obtained by selecting 'Manually configure' when your backup job is using a recommended filter, then filtering to see only 'Not recommended objects'.
Why everything isn't backed up by default
Perhaps counterintuitively, backing up every last Salesforce object in order to avoid any data loss isn’t the best approach. That’s because some objects tend to change frequently and provide little value when backed up.
For example, objects like AuthSession, LoginGeo and LoginIp store auditing data about login activity related to your Salesforce organization, so they see a lot of churn and don’t contain business-critical data.
Similarly, if you’re backing up a custom object like Case, you probably won’t get much benefit from backing up CaseFeed, CaseHistory, and CaseShare, which represent interactions with, field history, and sharing history of the base Case object. Gearset itself will capture the history of your objects, and so backing up Salesforce's history will likely provide little benefit for your business.
Finally, some objects (such as 'audit objects'; Log, Metrics, Usage), have data which is non-restorable into Salesforce, so having this in a backup may provide little value.
Not backing these objects up can provide clear benefits for your backup and Salesforce org.
As we will retrieve fewer objects, your backup jobs will run more quickly and use fewer Salesforce API calls.
For larger orgs, this may enable Gearset to run your backup more frequently, providing you with more granular backups and capture more states of your org.
By removing high-churn objects from your backup, anomalous changes to your org will become more obvious when viewing your 'job history'; this can help to spot issues with your Salesforce organization early, allowing you to rectify these issues sooner and with less disruption to your users.