The “leap second” of June 30 also caused problems on the network
A “leap second” (Leap Second or leap second according to Wikipedia ) is an extra second is added to atomic clocks in the world at certain times, so as to keep them synchronized and aligned with the rotational movement of the earth. Was first introduced in 1972 and to date over 20 who have been added in total at a rate quite irregular.
Yesterday June 30 added a new leap second to the list, exactly at 23:59:59 by Greenwich time, which marked the Time clocks you see in the picture above. This caused a little trouble on multiple servers scattered around the net and saw their associated services were offline for several hours, and of course, the case was the second extra time. The situation is quite simple: the Network Time Protocol is the protocol used to align the clock with atomic clocks (used in Java application platform or on the same Linux), and when you add the extra second to the clock, the NTP just gets to run in circles, hands on head, not knowing exactly what to do.
Among those affected by the second problem found Gawker Media, Reddit, StumbleUpon, LinkedIn, FourSquare and to Mozilla, all dependent Java, Linux, or any system using such a protocol NTP.
However, others like Google anticipated the event and prepared the NTP for adding milliseconds off the clock gradually and after the arrival of this second leap. Thus, there are no second repeat or anything like that and the time to Google’s servers still running as if nothing had happened.
The question is inevitable: what did they do yesterday with your extra second? Surely many wanted to use it given more time navigating the world of the internet, but among the problems generated by this and by the fall of the Amazon cloud , photos of the breakfast / lunch through Instagram or check-in at some favorite place had to wait.
Link: Why The Internet Died Again Last Night … (Gizmodo)Tags: atomic clocks, Java, leap second, Linux, Network Time Protocol, technical failure, Time