The alert is triggered when a host has fewer than 2 physical disks containing the system partition, resulting in insufficient system partition redundancy.
Host { .labels.hostname }: The system partition redundancy is not enough.
SCVM { .labels.hostname }: The system partition redundancy is not enough.
When ACOS is installed on a software RAID 1, two physical disks are used to build RAID 1 to store the system partition. When one of the system disks fails, the system automatically switches to the other system disk without affecting workloads, while the security level of the system decreases. If the other system disk fails as well, the node will not be able to provide any services.
A physical disk containing the system partition has failed, resulting in insufficient system partition redundancy on the host where the disk resides.
After logging in to the host:
Run lsblk to list all md devices. Common device names include md0, md126, and md127. The following commands use md127 as an example.
Run mdadm --detail /dev/md127 to check the number and status of member disks in the md device.
Run ls -lah /sys/block/md127/md/ to check the number of member disks under the corresponding md device path.
If the mdadm output shows fewer than 2 member disks, or any member disk is in the faulty state, the system partition redundancy is insufficient.
In AOC, unmount the affected physical disk, and then remount an SSD physical disk and select a cache disk with a metadata partition or a data disk with a metadata partition to create the system partition. This ensures the host has 2 physical disks containing the system partition.
