Reference Guide

Ungraceful system shutdown or power loss
If the host system experiences a power loss, the NVMe PCIe SSD may not have time to perform its internal shut down procedure. In such
an event, the device may enter a recovery mode.
This recovery process is also known as rebuilding. During rebuilding, there is very limited access from the host operating system. After the
recovery procedure is complete, the device is fully accessible from the host operating system.
NOTE: Dell recommends that you use power backup solutions for all Dell systems.
General errors
The following section describes general errors related to NVMe PCIe SSD.
NVMe drive properties intermittently not available in
iDRAC
Description NVMe drive properties via sideband (iDRAC) may not be available after a PCIe SSD is hot-inserted into the
system. This is most likely to occur if the PCIe SSD is formatted with a file system or has existing data.
Cause Side band controller on the NVMe drives does not complete initialization in time for iDRAC to inventory the
device.
Solution After an AC power cycle, the system should list the inserted devices in iDRAC.
NVMe PCIe SSD is not listed in the operating system
Cause
Hardware is not correctly installed.
Solution Check the following components:
• Devices: Ensure that the NVMe PCIe SSDs are installed in an NVMe PCIe SSD backplane.
NOTE: NVMe PCIe SSDs must be used with NVMe PCIe SSD backplanes. To ensure that you
have the correct configuration for the NVMe PCIe SSD, see the platform-specific owner’s
manual at www.dell.com/manuals.
• Backplane: Ensure that the cables for the NVMe PCIe SSD backplane are connected correctly.
• Cables: PCIe cables are unique for the configuration. Ensure that the backplane cable connectors are
connected to the backplane and the extender card or system board.
• Extender card: Ensure that the PCIe extender card, if used in your server configuration, is plugged into the
correct supported slot. See the system-specific owner's manual at www.dell.com/manuals.
I/O device error on write to NVMe PCIe SSD
Description
Windows event log may report the following entries on the first write attempt to an NVMe PCIe SSD: Event ID
7: The device, \Device\Harddisk\DRX, has a bad block.
When attempting to initialize the device using Computer Management > Storage > Disk Management, the
following message is displayed: Virtual Disk Manager, Data Error (cyclic redundancy
check).
Linux messages log may report the following entries on a write attempt to an NVMe PCIe SSD:
• Buffer I/O error on device nvmeXn1, logical block Y (where X is the number
corresponding to the device and Y is the logical block)
• nvmeXn1: unable to read partition table (where X is the number corresponding
to the device)
10 Troubleshooting