Diagnostics for VMware Cloud Foundation: ESX Health: SSD Storage Health

Products

• VCF Operations

Issue/Introduction

SSD (Solid State Drive): A persistent storage device utilizing NAND flash memory. Unlike traditional HDDs, SSDs have no moving parts, offering significantly higher IOPS and durability.

NVMe (Non-Volatile Memory Express): A high-performance communication protocol specifically built for SSDs. It connects via the PCIe bus to bypass the bottlenecks of older SATA/SAS interfaces.

 

1. Standard SSD (SATA/SAS)

ESX monitors traditional SSD health by tracking physical wear indicators such as reallocated sectors, total power-on hours, and operating temperature.

 

Status	Definition	Technical Indicators
Good	Optimal Health	The drive is operating within all manufacturer-defined parameters.
Warning	Potential Failure	Indicates the drive is beginning to show signs of wear or sector instability.
Critical	Imminent Failure / Offline	Drive Degradation: Failure is expected shortly. Device Error: The disk is unresponsive, permanently unavailable, or has been disabled by a user.

 

2. NVMe (PCIe)

NVMe devices use a high-performance protocol designed for low latency. Because of their architecture, ESX monitors specific metadata related to NAND endurance and controller status.

 

Status	Definition	Technical Indicators
Good	Optimal Health	All media and controller health logs are clear.
Warning	At-Risk	Spare Space: Available spare NAND is below 30%. Temperature: Drive temperature has exceeded the optimal threshold. Reliability: General reliability issues detected; replacement is recommended. Data Protection: A temporary error occurred while backing up internal memory to permanent storage. Data integrity may be at risk during power loss.
Critical	Immediate Action Required	Read-Only Mode: Media is locked to prevent data loss; no writes possible. Spare Space: Available spare NAND is below 10%. Device Error: The controller is no longer processing I/O or is permanently unavailable.

Environment

VCF Operations 9.x

Resolution

1. Verify via CLI: Log in to the ESXi host and run the following to see the physical device state:

esxcli storage core device list

 

1. Check SMART stats:
1.  For NVMe:
   esxcli nvme device log smart get -A <adapter_id>
   
   
2. For SATA:

esxcli storage core device smart get -d <device_id>

Action: If a device reports Critical spare space or enters Read-Only mode, evacuate all data (migrate VMs) and replace the physical drive immediately.