NSX Compute Manager Status DOWN with CorfuDB TransactionAbortedException and Error Code 90008

Products

• VMware NSX

Issue/Introduction

• NSX Compute Manager status is reported as DOWN in the NSX Manager UI despite no recent vCenter credential or certificate changes.
• cm-inventory.log file has following warning:

Aborted Exception org.corfudb.runtime.exceptions.TransactionAbortedException: TX ABORT ... Conflict Stream = nsx$ComputeManagerStatusModelMsg | Cause = CONFLICT

• A manual compute manager configuration save attempt triggered error: 

The time set on compute manager is ahead of the time set on nsx by ### seconds. (error code: 90008) 

Environment

VMware NSX

Cause

Time synchronization drift between the NSX Manager and the vCenter Server exceeds the 60-second tolerance threshold, causing database write conflicts and session failures.

Resolution

1. Synchronize the clocks of all NSX Manager nodes and vCenter Server nodes using a consistent NTP source.
2. From the NSX Manager UI, go to System > Fabric > Compute Managers.
3. Select the affected vCenter, click Edit, and then Save without changes to re-initialize the connection.