HANA Database LOG Backup using Recovery Services Vault

Jay Kim 11 Reputation points
2022-09-07T11:06:25.847+00:00

I have HANA Database and it's backed up by RSV,
the LOG backup policy is to backup every 1 hour. so I presume it will be every 1AM 2AM 3AM etc

now. we have VM auto reboot setup for cost saving. my question is.
when VM is rebooted at 7:45. is RSV LOG backup will kick in and try to backup LOG at 7:45 rather than wait for scheduled time 8AM?

the reason why I'm asking this is, I am getting backup failure everyday 10 mins after reboot schedule 7:55 - 8:10

SAP HANA on Azure Large Instances
SAP HANA on Azure Large Instances
Microsoft branding terminology for an Azure offer to run HANA instances on SAP HANA hardware deployed in Large Instance stamps in different Azure regions.
120 questions
0 comments No comments
{count} votes

1 answer

Sort by: Most helpful
  1. Prrudram-MSFT 22,921 Reputation points
    2022-09-12T16:19:30.897+00:00

    Hello @Jay Kim ,

    Thank you for reaching out to the Microsoft Q&A platform. Happy to answer your question.

    As per your update, I understand you have specified the log backup frequency is set to every one hour in the backup policy like below.
    240167-image.png

    Please note that all the scheduled backups and on-demand backups (triggered from the Azure portal) that are either full or differential are initiated by the Azure Backup Plugin for HANA. However, log backups are managed and triggered by HANA Backup Engine itself.

    About your assumption of log backup schedule to be triggered sharp 7 am or 8 am is not correct. The backup intervals are always based on the start time/end time of the last successful backup. If the log backup was at the very first is triggered at say 7:30 or 7:45 am the next schedule of the log backup becomes 8:30 am or 8:45 am.

    In your case I see the log backup time is conflicting with your VMs reboot time. VM is shut down during the operation

    Azure backups have the logic to detect this in Fabric Service to get the VM running status and mark the operation a "User Error if the VM was found shut down. But, if the VM is shutting down during the operation, ARM keeps reporting the VM as running for some time and hence the operation fails with error similar to "WorkloadExtensionNotReachable". It is helpful to check the error message reported for the failed log backups and refer to the following document

    Suggestion: Based on the log backup start/end time in the backup job history, it will be good to re-adjust the VM reboot schedule by noting the total time taken to reboot the VM.

    --please don't forget to upvote and accept as answer if the reply is helpful--