Backup requests for ServiceNow instance including productionSummaryThis article provides information about ServiceNow instance backups. Unfortunately, we do not have an option for on-demand backups of any instance. We do not have the option to take sparse table backup and the automated backup process takes the backup of the whole instance Full (Weekly) and Incremental/Differential (Daily). There is no option to do a sparse table restore and any restore request will be for the whole instance with the snapshot selected for that day. Our backup policy is available in the following Knowledge Article: Instance Backup and Recovery. An on-demand backup for production instances is not a reliable/foolproof mechanism for protecting your PROD instance data in case of any major change/deployment/upgrade on your instance. An example scenario is as follows: A differential daily backup can take >10 minutes and a Full Weekly backup can take much more than that.If you have a deployment at 22:00, you request us to take a backup at 21:30Even if we take the backup at 21:30 and say it finishes at 21:40 and your instance is live and new data getting added till you start your deployment.There will be a lot of data changes happening on an instance even if it is at OOB hours your Schedule Jobs, PA scores, and Integrations Jobs might be configured for OOB hours.Deployment took 15 minutes and you decide it went bad and you need to restore the PROD with backup.Restoration will take close to 6 hours ( Ballpark estimate).If a restore is initiated at 22:30 and around 03:30 (Ballpark time) the restore is complete.That instance will be missing the DATA from 21:40 to 03:30 AM For the above-said reason, a PROD ad-hoc backup is not useful in any scenario. We have other standard processes in place by default to get a similar outcome in case you need a PROD data replica for any given time for the last 4 days. Your data is safe with us. In a most unlikely scenario of data corruption in production, we have an option to do a Point in Time (PIT) restore of the production instance to a temporary instance by which we can bring back the data on the PROD instance to that specific time for data comparison (to a specific minute) on the temporary instance. A point-in-time restore is done by restoring the instance to the last available backup and recreating the remaining data specified from the BIN logs/Transaction Logs. Auto retention of BIN LOGS is 4 days as of this writing. PIT restore is a manual process involving multiple teams and this will be initiated in Critical conditions and cannot be treated as a Failover/Recovery/Back-out/Rollback Method. Note: We do not do PIT restore directly on PROD and we will only fix forward and the PIT restore will be used only for Data comparison and Data recovery on TEMP/SUBPROD instances. Once you have the data on the TEMP instance, this can be used to compare the data between POST and PRE upgrade/deployment/change versions and you can work on a data recovery plan. We can assist with any roadblocks as we have standard operating procedures to transfer the data based on individual cases. More information on this process (PIT Restore) is covered in this documentation. Fix Forward vs Restore from BackupMore information on Backups and Retention is covered in this white paper (Page 17): Delivering Performance, Scalability, and Availability on The Servicenow Cloud This is applicable for any instance including non-production instances. Dataloss on sub production or non production instances KB has some scenarios listed and the options to recover the data.