ESXi host fails with PSOD "Fatal error" with Intel E810 vmnic.

Products

• VMware vSphere ESXi

Issue/Introduction

• Intel E810 vmnic with icen driver is installed on ESXi.
• ESXi Host fails with PSOD with "Fatal error"
• The scenarios above may lead to PSOD with the following information:

IPMI log:   Record:2:    Record Id: 2    When: YYYY-MM-DDTHH:MM:SS    Event Type: 111 (Unknown)    SEL Type: 2 (System Event)    Message: Assert + Critical Interrupt Bus Fatal Error    Sensor Number: 24    Raw:    Formatted-Raw: 02 00 02 3a 08 d3 63 b1 00 04 13 18 6f XX XX XX

• PCI list shows the PCI id relates to E810 vmnic:

vmnic10  0000:b1:00.0     Up    25000  Full    9000  icen    1.9.8.0         4.00 0x800118b3 21.5.9  b4:83:51:03:XX:XX  8086  159b  8086  0002  Intel(R) Ethernet Controller E810-XXV for SFP     vmnic11  0000:b1:00.1     Up    25000  Full    9000  icen    1.9.8.0         4.00 0x800118b3 21.5.9  b4:83:51:03:YY:YY  8086  159b  8086  0002  Intel(R) Ethernet Controller E810-XXV for SFP

 

• In some customer's environments, IPMI log shows as below

Record Id  When                 Event Type     Message       ---------  ----                 ----------     -------       292        YYYY-MM-DDT09:01:25  111 (Unknown)  Assert + Critical Interrupt Bus Uncorrectable error       293        YYYY-MM-DDT09:01:44  111 (Unknown)  Assert + System Event Undetermined system hardware failure       294        YYYY-MM-DDT09:04:48  111 (Unknown)  Assert + Slot/Connector Fault Status

 

Environment

VMware vSphere 7.0.x

Cause

'Critical Interrupt Bus Fatal/Uncorrectable error' is obviously a hardware issue, not VMware software part.

Resolution

Involving Intel hardware vendor with above symptoms is recommended.

Workaround:

No workaround.

Intel vendor is still investigating the issue. Pending on Intel's output.