Troubleshooting ESXi PSOD Fatal PCIe Hardware Error (vSphere ESXi)

Products

• VMware vSphere ESXi

Issue/Introduction

Hosts encountering a purple screen of death (PSOD) due to PCIe hardware errors often indicate a failing network interface card or root port. This article provides technical guidance to diagnose and resolve fatal hardware interrupts related to Advanced Error Reporting (AER) and PCIe subsystems in vSphere environments.

• Host becomes unresponsive or displays a PSOD.
• System logs report ApeiHEST fatal error originating from a specific PCIe device.
• Logs contain AERUeSts: 0x20 (Advanced Error Reporting) or LINT1/NMI entries.

Environment

• VMware vSphere ESXi 8.x
• Physical Server Hardware (Dell/HPE)

Cause

The fatal error is triggered when the ESXi kernel detects an uncorrectable PCIe error reported by the hardware bus (e.g., PCI Express Root Port). This often stems from a physical failure of the Network Interface Card (NIC) or the associated PCIe slot, forcing the host to halt to prevent data corruption.

Resolution

1. Extract Diagnostic Logs: Access the host console or extract the support bundle via vCenter to examine vmkernel.log. Search for ApeiHEST and AERUeSts to pinpoint the failing device address.
2. Verify Hardware Status: Log in to the host's BMC (iDRAC for Dell, iLO for HPE) and check the system event logs for hardware-level PCIe errors or slot faults.
3. Physical Inspection: Place the affected host in maintenance mode. Power down the server and physically inspect the NIC and its PCIe riser. Reseat the card and ensure cables are securely connected.
4. Hardware Replacement: If the fatal error persists after reseating, replace the network interface card. Monitor the host after replacement to confirm the AER errors cease.
5. Validate Initialization: Once hardware is replaced, boot the host and run esxcli hardware pci list to confirm the new NIC is detected and initialized correctly.

Additional Information

To speak with a customer representative or a Support Engineer, see Contact Support. Scroll to the bottom of the page and click on your respective region.