This topic explains how to set up Virtual RAID on CPU (VROC) on Generation 6 hardware where a failed drive has been replaced.
Use the output from the following command to find the volume with the failed drive. Drives with
[UU]are working normally, and drives with[U_UUUU]have failed. In this example,md123is the failed drive.cat /proc/mdstat Personalities : [raid0] [raid1] [raid6] [raid5] [raid4] md122 : active (auto-read-only) raid1 nvme4n1[1] nvme5n1[0] 937687040 blocks super external:/md125/0 [2/2] [UU] md123 : active raid5 nvme2n1[4] nvme7n1[3] nvme0n1[2] nvme3n1[1] nvme6n1[0] 31256089600 blocks super external:/md126/0 level 5, 128k chunk, algorithm 0 [6/5] [U_UUUU] md124 : active (auto-read-only) raid1 nvme8n1[1] nvme9n1[0] 6251217920 blocks super external:/md127/0 [2/2] [UU] md125 : inactive nvme4n1[1](S) nvme5n1[0](S) 10402 blocks super external:imsm md126 : inactive nvme2n1[4](S) nvme6n1[3](S) nvme3n1[2](S) nvme0n1[1](S) nvme7n1[0](S) 26005 blocks super external:imsm md127 : inactive nvme8n1[1](S) nvme9n1[0](S) 10402 blocks super external:imsmIn the output details for
md123, find the parent volume. This is the volume name afterexternal:. In this example, the parent volume ismd126.Use the following command and note which device name is in the output, but is not listed in any volume of the
cat /proc/mdstatoutput. In this example, it isnvme1n1.lsblk --nodeps -o name,model,serial NAME MODEL SERIAL nvme3n1 Micron_7450_MTFDKCC6T4TFS 23084033D60C nvme6n1 Micron_7450_MTFDKCC6T4TFS 23084033D74E nvme7n1 Micron_7450_MTFDKCC6T4TFS 23084033D574 nvme1n1 MTFDKCC6T4TGQ-1BK1JABYY 052510B7CF53 nvme2n1 Micron_7450_MTFDKCC6T4TFS 23084033D60B nvme4n1 Micron_7450_MTFDKBA960TFR 23053F109F15 nvme5n1 Micron_7450_MTFDKBA960TFR 23053F109F4F nvme9n1 Micron_7450_MTFDKCC6T4TFS 23084033D451 nvme0n1 Micron_7450_MTFDKCC6T4TFS 23084033F771 nvme8n1 Micron_7450_MTFDKCC6T4TFS 23174039A362Use the following command to add the replacement drive, including the metadata volume name and the NVMe drive device name:
mdadm --manage /dev/md126 --add /dev/nvme1n1Use the following command to monitor the rebuild:
cat /proc/mdstat(Optional) Enable the maximum rebuild rate, effective until the next reboot:
echo 99999999 > /proc/sys/dev/raid/speed_limit_max echo 99999999 > /proc/sys/dev/raid/speed_limit_min