A company has moved all on-premises workloads into a public cloud. After some time, the cloud engineer starts noticing time drifts on the VMs and suspects an NTP issue. Time drifts were not an issue…

Implementing UTC across all cloud workloads establishes a single, universal time standard that is independent of geographic region or time zone configuration, eliminating drift caused by inconsistent time settings across distributed VMs. On-premises environments often rely on a domain controller or local hardware clock as the authoritative NTP source, but in cloud environments, UTC is the universal reference and ensures all VMs share a consistent time base. Standardizing to UTC resolves drift that arises when VMs inherit conflicting or incorrect time zone settings during cloud provisioning.

Routing cloud VM time synchronization through an on-premises NTP server introduces network latency, creates a dependency on the corporate WAN, and establishes a single point of failure for all cloud workload time synchronization.

Pointing workloads to a cloud provider stratum 2 server addresses the NTP source but does not resolve drift caused by inconsistent time zone or time standard configurations across the VMs themselves.

Configuring VMs as NTP servers is unreliable because VM clocks are themselves subject to drift and virtualization clock skew, making them unsuitable as authoritative time sources for other workloads.