Timestamp Drift Checker

Verify NTP service is running and synced

Run timedatectl (Linux) or w32tm /query /status (Windows). Check that the system clock shows the correct time with an active NTP sync.

Check current clock offset from UTC

Use timedatectl timesync-status (Linux) or w32tm /stripchart (Windows). Offset should be under 100ms.

Ensure server timezone is set to UTC

Exchanges expect timestamps in UTC. Run timedatectl set-timezone UTC (Linux) or set via control panel (Windows).

Use an appropriate recvWindow value

Start with the exchange default. Increase if you consistently get timestamp errors. Lower values reduce risk but increase errors.

Fetch server time before signing

Call the exchange time endpoint before signing each request. Calculate offset = serverTime - localTime and adjust your timestamp.

Check if running on a VM with clock drift

Cloud VMs (AWS, GCP, Azure) are prone to clock drift. Use NTP with multiple servers and monitor offset regularly.

Set up clock drift monitoring

Alert when drift exceeds 500ms. Track drift over time to identify patterns (e.g. spikes after VM resume).

Implement fallback time sync strategy

If NTP fails, implement a fallback: fetch exchange server time, compute drift, and adjust timestamp before signing.