vSphere system with EPYC processors of the series 7002, 7Fx2, 7Hx2, and 7001 might become unresponsive after more than 1000 days of continuous uptime

Products

• VMware vSphere ESXi 8.0

Issue/Introduction

vSphere system with EPYC processors of the series 7002, 7Fx2, 7Hx2, and 7001 might become unresponsive after more than 1000 days of continuous uptime

 

Environment

Esxi 8.x

Cause

Due to AMD erratum 1474, if the core-C6 (CC6) sleep state is active on your EPYC processors of the series 7002, 7Fx2, 7Hx2, and 7001, a core might fail to exit CC6 in about 1044 days after the last system hardware reset, which excludes reboots by using VMware QuickBoot.

As a result, your vSphere system might become unresponsive.

Resolution

This issue is resolved in ESXi-8.0U3. The fix automatically deactivates CC6 until the next hardware reset.