ESXi host reports "No healthy upstream" error after enabling Hyper-threading
Article ID: 436916
Updated On:
Products
Issue/Introduction
After enabling Hyper-threading on an ESXi host via hardware management, the host may report a "No healthy upstream" error, leading to an environment down state or network instability. This typically manifests as a failure of the host to properly communicate with the Distributed Virtual Switch (DVS) or upstream network components following the hardware change.
Environment
- VMware ESXi 8.x
- Distributed Virtual Switch (DVS)
Cause
The issue is caused by a synchronization failure between the host's physical hardware state and the virtual networking configuration (DVS) after a low-level hardware modification (Hyper-threading). This can result in the host losing valid upstream network paths.
Resolution
To resolve this issue, perform a clean reconfiguration of the host's hardware and storage states:
Workload Migration: Migrate all active Virtual Machines (VMs) from the impacted host to other available hosts in the cluster by placing the host in maintenance mode.
Storage Preparation: Temporarily disconnect LUNs from the ESXi host at the storage array level.
Hardware Optimization: Apply the updated Hyper-threading settings via the UCS Manager.
Host Reboot: Reboot the ESXi host to commit the hardware changes.
Network Verification: Verify Distributed Virtual Switch (DVS) settings to ensure network stability and compare the DVS settings with healthy host.
Storage Restoration: Reconnect the LUNs from the storage end and perform a Full Storage Rescan on the host to ensure all datastores are accessible.
- Validation: Exit Maintenance Mode and migrate a test VM back to the host to verify network connectivity and performance.
Additional Information
Feedback
