CONTAP-729114: FSx for ONTAP HA pair can get stuck in takeover/reboot loop after throughput scaling when takeover-with-halt overlaps with check-takeover
Issue
- In FSx for ONTAP (AWS), after a throughput scaling operation (instance type change) and a planned/negotiated takeover, if a takeover-with-halt occurs while check-takeover runs (often triggered by Upgrade Preparedness Check), the rebooting node may fail to complete boot/giveback and repeatedly halt/reboot.
- Node01 is experiencing a boot loop due to a 'serial number not found' error:
ena1: <ENA adapter>SN(volxxx) does not match ephemeral nvram bootarg.vm.vnvram.sn=AWSxxx disk_ident(volxx mem 82e)
Root mount waiting for:SN(volxxx) does not match ephemeral nvram bootarg.vm.vnvram.sn=AWSxx disk_ident(volxx CAMena1: ENA Express tuning is enabled.
Ephemeral vNVRAM SN (AWSxxx) not found, clear bootarg.vm.vnvram.sn and reboot
SerialNumber(AWSxxx) matches ephemeral vNVRAM, using bootarg.vm.nvramdevice=/dev/nda8
- HA remains degraded until manual recovery.
