ESX hosts intermittently going into a "not responding" state on the vSphere UI

Products

• VMware vCenter Server

Issue/Introduction

The impacted ESX hosts may have been added to the environment recently.

The vCenter is configured with both a primary and secondary DNS server.

In the vpxd log on vCenter, missed heartbeats are observed in /var/log/vmware/vpxd/vpxd.log:

Missed ## heartbeats for host [vim. HostSystem:host-###, <esxi_host_name>

However, heartbeats have been confirmed to have been sent by the host and received by the vCenter by doing a packet capture using https://knowledge.broadcom.com/external/article?articleNumber=307364

 

Environment

VCF 5.x

Cause

The secondary DNS server has incorrect PTR (Pointer) records (IP address to fqdn) for the impacted ESX hosts.

This can be confirmed by doing an nslookup on the vCenter to the impacted ESX hosts fqdn - it will be observed that when the host shows as "not responding" in the vSphere UI, the nslookup returns an unexpected IP address for the ESX host.

Resolution

Correct the incorrect PTR (Pointer) records (IP address to fqdn) for the impacted ESX hosts on the secondary DNS server.