NTP Operational Thresholds

Products

• VMware vSphere ESXi

Issue/Introduction

Time syncronization issues on ESXi can appear even when NTP is properly configured. To ensure functionality, the thresholds below should not be exceeded. Any values exceeded should be investigated for configuration or upstream issues.

Environment

VMware ESXi Server 7.x
VMware ESXi Server 8.x

Cause

Possible causes for this can be following but not limited to. 

• Connectivity between NTP server and ESXi.
• Configuration Issues.
• Firewall related issues.

Resolution

Ensure the following thresholds are not breeched for the values mentioned:

ESXi NTP Operational ThresholdsStratum (Source Hierarchy)

• Healthy: 1–3.
• Warning: 4 or higher. While functional, increasing distance from the root clock increases the risk of error.
• Critical: 16. This is the protocol limit; at this point, the host considers the source invalid and will not synchronize.

Root Dispersion (Maximum Distance)

• Healthy: < 1000 ms.
• Warning: 1000 ms – 1499 ms.
• Critical: 1500 ms (1.5 seconds). This is a hard-coded threshold in ESXi. If dispersion reaches this value, the host will reject the time source.
• Note: If your environment frequently nears this limit, the tos maxdist parameter can be used to increase tolerance, but the default is 1.5s.

Reach (Connection Success)

• Healthy: 377. This indicates that the last 8 consecutive poll attempts were successful.
• Warning: Any value less than 377. This indicates intermittent packet loss or failed polls.
• Critical: 0. The server is completely unreachable or not responding to NTP queries.
• Clock Offset (Time Difference)

Clock Offset (Time Difference)

• Healthy: Values close to 0 ms.
• Warning: > 50 ms. Large offsets can indicate underlying network latency or server instability.
• Critical: 128 ms. When the offset exceeds this value, the ESXi NTP daemon will perform a step-correction (forcing the clock to jump) rather than a gradual slew. This can be disruptive to applications sensitive to time continuity.

Jitter (Latency Variation)

• Healthy: < 10 ms.
• Warning: Significant spikes in jitter often precede increases in dispersion and offset, serving as an early indicator of network congestion or NTP server load.

Additional Information

Troubleshooting NTP on ESX and the ESXi 7.x / 8.x.

'WARNING: NTPClock: 680: system clock apparently no longer synchronized to upstream time servers' on ESXi