Leap Smear

Overview

Leap smear is a method for handling leap seconds without introducing the normally non-existent timestamp 23:59:60. Instead of inserting a sudden extra second, the adjustment is spread across a window of time by making each second very slightly longer. Google pioneered this approach in 2008, and it has since been adopted by major cloud providers including Amazon (AWS) and Microsoft (Azure).

How It Works

In Google's implementation, each second within a 24-hour window (12 hours before and after the leap second) is lengthened by 1/86,400 of a second, approximately 11.6 microseconds. Over 86,400 seconds, these tiny additions sum to exactly one full second. During the smear window the system clock drifts up to 0.5 seconds from true UTC, but at no point does an anomalous 23:59:60 timestamp appear.

Provider Differences and Caveats

The start time and duration of the smear window differ among providers. Google and AWS both use a 24-hour smear, though their start times may not be identical. Some providers use shorter windows of 1 or 2 hours. When comparing timestamps across different cloud environments around a leap second event, discrepancies of up to one second can occur due to these differing smear schedules. After the planned abolition of leap seconds in 2035, leap smearing will become unnecessary.