Linux blamed for 'leap second' that humbled Internet - Page 1 - Enterprise Infrastructure

[Jamon Camisso]

On 12-07-03 01:54 PM, Kevin Cozens wrote:
> On 12-07-03 01:31 PM, William Park wrote:
>> They blame the kernel, but only software affected are Java-based. :-)
> 
> So why where people reporting issues with MySQL?
> 
> I find the whole thing a bit puzzling. I've know about leap seconds for
> years. They aren't new or even that rare as they seem to occur on a
> regular basis. I've heard quite a few of them when listening to WWV on
> shortwave. Why did this recent one cause so many headaches.

I believe google setup their ntp servers to gradually add a few
milliseconds over the days leading up to the leap second and as such
weren't affected.

https://lkml.org/lkml/2012/7/1/203 seems to be the root cause analysis,
seems like it has been around since 2.6.22 kernel.

Running date -s "$(date)" was enough to fix it - just needed to give the
clock a kick is all. A reboot worked too.

Jamon
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists