SD-WAN Edge service failure after link flapping with SNMP in HA

Products

• VMware
• VMware SD-WAN by VeloCloud

Issue/Introduction

Symptoms:
The VCO Events will show some events such as the below:

• HA failover (High Availability Going Active)
• One or more consecutive interface flapping within the same time (Edge Interface UP/Edge Interface down/Link Alive/Link Dead).

• User may also notice an event with a message "Edge dataplane service failed"

Environment

VMware SD-WAN by VeloCloud
VMware SD-WAN

Cause

The issue is identified with ID 95565 and it is triggered by the HA Edge's WAN links flapping one or more times (going down and then come up rapidly) while also using SNMP where there are frequent SNMP queries. There is a timing issue where the interface coming back up and the SNMP query together can trigger a Dataplane Service failure and cause a failover. While only a single WAN link flap can cause this issue, the greater the frequency of WAN link flaps, the greater the potential for this issue to occur.

Resolution

This issue is resolved in the versions below It does not affect version 4.3.x or below. For information on how to upgrade please check the following article: VMware SD-WAN Software Upgrade FAQs
4.5.1      R451-20220916-GA  
5.1.0      R5100-20221204-GA

Workaround:
Disable SNMP