NSX Distributed Load Balancer Shows Degraded or Partially Ready Status with Not Ready LSPs

Products

• VMware NSX

Issue/Introduction

In NSX environments where Distributed Firewall is enabled and no workloads are present in the DFW exclusion list, the Distributed Load Balancer status may report as degraded in the user interface.

Checking the DLB status via the command line interface reveals a partially ready status with a non-zero Not Ready LSP Count

 

> get load-balancer <REDACTED_UUID> status
Tue Mar 17 2026 UTC 11:30:59.104
Load Balancer
UUID               : <REDACTED_UUID>
Display-Name       : clusterip-domain-<REDACTED_STRING>
Status             : partially ready
Ready LSP Count    : 169
Not Ready LSP Count: 13
Conflict LSP Count : 0
Ready LSP          : <REDACTED_UUID>
                     <REDACTED_UUID>

Not Ready LSP      : <REDACTED_UUID>
                     <REDACTED_UUID>
                     <REDACTED_UUID>
                     <REDACTED_UUID>

Environment

VMware NSX 

VCF NSX 9.0 

Cause

This is an intermediate state caused by asynchronous multithreading. The Load Balancer and Logical Switch Port  are configured in different threads, causing DLB and LSP configuration messages to reach the datapath at different intervals. The DLB status is polled and reported by the aggregation service every minute. In setups where DLB and LSPs are reconfigured frequently, this transient "not ready" status is highly likely to be captured and surfaced as degraded in the UI. This is a reporting artifact and does not impact datapath traffic.

Resolution

Acknowledge that the degraded UI status is a reporting artifact and safely ignore the warning if datapath connectivity remains unaffected.

Upgrade the environment to VCF NSX 9.1. In VCF NSX 9.1, the DLB architecture has been separated from the vsip module and dvfilter, permanently resolving the "LSP not ready" reporting condition.