Skip to content

Mismatched log timestamps: clock drift or a time-zone issue?

Read this article in ไทย 中文

Mismatched log timestamps: clock drift or a time-zone issue?

Do not change clocks just to tidy a report

An IT team sees a seven-hour difference between two logs and assumes a machine has the wrong time. First check whether one records UTC while the other displays Thai time. In a hypothetical example, 03:00 UTC and 10:00 in Thailand can represent the same event with both clocks correct. Moving a machine back seven hours would create a new problem.

A different case is actual clock drift despite identical display zones. A log collector may also receive a record later than the event occurred. These are separate issues: changing a time zone does not fix clock drift, and resetting a clock does not clear a delayed log queue.

Map timestamps from source to report

Choose a traceable sample, such as an internal test with a shared transaction identifier. Record event time, receipt time and report display time, including each field’s time zone or offset. Identify which component creates each timestamp: application, database or collector. Screen order alone is not evidence of event sequence.

If a log omits its zone, retain the original format and obtain configuration evidence rather than guessing UTC or Thai time retrospectively. Normalize timestamps in a separate analysis copy while preserving originals, and record the conversion rule so another reviewer can reproduce the comparison.

Inspect the time source in its domain context

Microsoft explains that Windows Time uses NTP to synchronize clocks and that synchronization matters to services and Kerberos authentication. Most domain members follow the domain time hierarchy; the forest-root PDC emulator typically synchronizes with an external source. Do not point every computer at one external server without checking the system design.

Initial checks should record the active source, latest synchronization status, measured offset and clock-change events. For a virtual machine, have the administrators review guest and host settings together. Separate measured evidence from hypotheses such as network failure or a stopped service before choosing a fix.

Preserve evidence and control the change

Before changing settings, preserve raw logs and existing configuration with the investigation start time. For a security incident, let the incident lead define evidence handling; never rewrite original timestamps to make the narrative agree. Production clock changes need an approver, a change window and a plan to monitor time-dependent applications.

After the change, repeat the traceable event through the complete source-to-report path. Check both timestamps and identifiers, and watch for recurring drift. Matching screen clocks once does not establish continuing synchronization or validate historical logs created before the fix.

Set acceptance criteria around the workload

When asking Chawinnix to support IT or security systems, prepare a system map, sanitized log samples and the questions the timeline must answer. Agree acceptable timing error with the workload owner. Do not turn a manufacturer’s accuracy statement into a guarantee without its conditions. The aim is traceable events and explicit evidence limitations.

Server blocks with clocks connected to a central reference clock

Sources

Back to all articles

Start the conversation

Tell us briefly what concerns you and our team will get back to you. What you send stays confidential.