Partner performed FFU 16.2 to 17.1. After performing the upgrade partner noticed that the cinder backup service stopped working and it was not possible to create new backups. After investigation it was found that a recursive chgrp on the backup NFS share was done for each backup, which is taking around 2 hours blocking the backup service. NFS share is mounted inside the backup container.

This however was working fine with the same share on 16.2, so it seems like a regression. Customer Portal case: 04115559

To Reproduce Steps to reproduce the behavior:

1. Create a backup
2. See chgrp process running
3. No backup possible until the chgrp is finished

Expected behavior

• Backup should work as before the FFU

Device Info (please complete the following information):

• NFS share is served by Netapp

Bug impact

• The partner is a cloud provider offering self-serviced cloud environment to their end customers. All end customers are not able to perform backups. This has significant impact on the partner.

Known workaround

• Possible workaround could be to try to add NFS options to cinder conf