A few days ago, a customer asked me about the modalities for ServiceNow instance back-up and restore process.

It’s a topic I have (fortunately) not come across so often along my career. I had to re-do my research every time.

Here is a summary of the key information.

PRODNon-PROD
Differential backupDailyDaily
Full database backupWeekly Weeky
Retention period14 days (can be extended to 28 at a cost)14 days
On-demand backup optionNoNo

How to request an instance restore

The restore process is done from the ServiceNow support portal:

When to restore and when to clone

In the back-up request form, there is a choice for “Source instance” and “Target instance”. It looks pretty much like a clone.

Except that with a clone you don’t have choice of which snapshot of the source instance will be taken (usually a couple of hours before the time the clone is executed)

Besides, using the back-up from an instance to another works around features like table exclusions and Data Anonymization.

There may be use cases where Restore of a previous instance version over another instance is relevant, but that’s probably not too common.

When I have seen a PROD restore

Surveying my customers… more on that coming.

  • During first ITSM implementation, a script was ran to delete all Tasks from PROD, to have a clean instance ready for go-live. It was only forgotten that the STORY table extended from the TASK table so stories were lost in the process. The only way to recover it was to use a Restore (Today, you can also do it one by one by going in the “Deleted items table).

Considerations:

Restoring an instance is a last resort option, due to the loss of data (all records created since the back-up date), the fact that it’s “all or nothing”. and the PROD unavailability during the process.

Leave a comment

Trending