Stale networks (NSX segments) are observed in the vCenter

Products

• VMware NSX Networking
• VMware NSX

Issue/Introduction

• vCenter is used as a Compute Manager in NSX deployment.
• In Networking tab in the vCenter, NSX networks which have been deleted in NSX Manager, are still present.
• On NSX Manager, logical segments are in "failed" state in GET https://<nsx-manager-ip>/api/v1/logical-switches/state.

Environment

NSX 4.x

Cause

During segment deletion, intermittent issues (e.g. connectivity issue) may cause logical segment to get to failed state.

Logical switch auto rectification mechanism (NSX Manager's internal process) does not work for logical switches in failed state, hence auto rectification mechanism may fail to automatically rectify logical switches which are in FAILED state.

Due to this, port groups will not get cleared from the vCenter. 

Resolution

This issue has been resolved in NSX 4.1.2.5, and NSX 4.2.0 and higher. 

Please refer to VMware NSX 4.1.2.5 Release Notes, "Fixed Issue 3413790" discussed in section "Resolved Issues".

 

Workaround:

To workaround this issue, stale/orphan port groups must be manually deleted from the vCenter's database. For assistance with this task, please open a new case with Global Support team with reference to this KB article.