Nvidia DGX-2 System Service Manual page 14

Identifying the Failed NVMe from the Console
To identify the failed NVMe drive from the DGX-2 console, enter the following and then look for
a missing entry from the output.
sudo mdadm -D /dev/md1
$
Number Major Minor
0 259
1 259
2 259
3 259
4 259
5 259
6 259
7 259
The list should include device names from
drives, and from
nvme0n1
To map the device name to the physical slot ID, enter the following, where X corresponds to
the missing device name.
ls -l /dev/disk/by-path |grep nvmeX |cut -d'|' -f3
$
The command returns the PCIe bus ID. Refer to the following figure to find the slot ID that
corresponds to the PCIe bus ID for the faulty drive.
DGX-2 System
RaidDevice State
8 0 active sync
13 1 active sync
7 2 active sync
10 3 active sync
12 4 active sync
11 5 active sync
9 6 active sync
6 7 active sync
nvme2n1
through for systems with 16 NVMe drives.
nvme15n1
U.2 NVMe Cache Drive Replacement
/dev/nvme9n1
/dev/nvme5n1
/dev/nvme6n1
/dev/nvme3n1
/dev/nvme2n1
/dev/nvme7n1
/dev/nvme8n1
/dev/nvme4n1
through for systems with 8 NVMe
nvme9n1
DU-09224-001 _v09 | 8