Anyone familiar with vCenter services/feature should have heard about dnsmasq, VCSA has this service installed and started by default which is introduced from 6.0.

DNSMASQ service is a internal DNS service that aggressively cache lookups far beyond the TTL of the A record.

If you are given a task of changing the IP address of few ESXi servers in a vSphere cluster to a new network segment. Generally you start with making the changes to the ip in ESXi host , update DNS and re-add the hosts back to vCenter inventory using FQDN. But this might not well if the vCenter has the dnsmasq service started, where you might end up seeing hosts in disconnected state or cannot add the hosts to vCenter server with FQDN – can be added using ip though. In this scenario, restart of dnsmasq service will help you to fix the issue.

Check the dnsmasq service status in VCSA using below command: 

systemctl status dnsmasq

Restart the dnsmasq service using below command and this will not have any ill effects on the infrastructure, so safe to do anytime.

systemctl restart dnsmasq

The dnsmasq service holds hostname records in a text file on the vCenter Server Appliance /var/lib/vmware/dnsmasq/hosts.

One issue we are seeing with dnsmasq is the root partition on the vCenter server appliance filling up due to dnsmasq.log files: work around is available here