Health Monitors are getting removed even though "Health Monitors to be considered "Up" is greater than 0.

Products

• VMware Avi Load Balancer

Issue/Introduction

• We observed a behavior where Health Monitors (HMs) are being removed from a pool even though the "Health Monitors to be considered Up" setting is configured to 2.
• This occurs specifically when a server is disabled via VMware Cloud Director (VCD). Due to a known bug on the VCD side, the pool configuration is updated in a way that unintentionally removes the associated Health Monitors.
• Pool will do down due to this Error: "State from remote SE/Unknown failure"

Cause

In earlier Avi LB releases (prior to v31.1.1), the system skipped validation when a Pool was configured with:

• No health monitors attached, and

• min_health_monitors_up set to a value greater than 0

This behavior led to pools being marked DOWN, even though no health monitor was present to evaluate the health of pool members.

 

Resolution

Resolution

This logic has been removed starting with version 31.1.1, and now the validation correctly checks for such mismatches.

• If min_health_monitors_up is set to a value greater than 0,

• At least one health monitor must be attached to the pool.

Otherwise, a validation error will be raised during configuration.

 

Recommendation

• If you're running a version older than 31.1.1, ensure that:

  • If min_health_monitors_up is set, attach at least one health monitor.

  • Or, set min_health_monitors_up = 0 to avoid false DOWN state.

• Upgrade to 31.1.1 or later to benefit from improved validation and avoid silent misconfigurations.