[LEAPSECS] postmortem, specifics of fails

[Steve Allen]

It's nice to see professional journalists chasing the stories of fail
http://www.wired.com/wiredenterprise/2012/07/leap-second-bug-wreaks-havoc-with-java-linux/

Things that happened before the leap second

a linux kernel bug was noted in March
https://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=6b43ae8a619d17c4935c3320d2ef9e92bdeed05d
reports indicate that not all vendors implemented this fix

Redhat initially believed that RHEL6 was not affected, but newer ones were
I'm not sure if the complaints about JVMs are disambiguated from kernels.
https://access.redhat.com/knowledge/articles/15145
But they later discovered that running NTP could cause a livelock
before the leap second
https://access.redhat.com/knowledge/solutions/154713
and their fix seems to be
https://access.redhat.com/knowledge/articles/15145

The reports about Debian squeeze started well before the leap second
and are also related to the NTP leap flag
http://serverfault.com/questions/403732/anyone-else-experiencing-high-rates-of-linux-server-crashes-during-a-leap-second

Things that happened at the leap second

java virtual machines and mysql seemed unhappy.
It isn't clear whether it was one of those or the two together.
They were fixed by setting the system date to the current date.
http://blog.mozilla.org/it/2012/06/30/mysql-and-the-leap-second-high-cpu-and-the-fix/

There were also reports of problems with ruby, but I have not seen
details of the kernel version, NTP config, or coexistence of mysql.

linux system logs showed issues with futex
https://lkml.org/lkml/2012/6/30/122

--
Steve Allen                 <sla at ucolick.org>                WGS-84 (GPS)
UCO/Lick Observatory--ISB   Natural Sciences II, Room 165    Lat  +36.99855
1156 High Street            Voice: +1 831 459 3046           Lng -122.06015
Santa Cruz, CA 95064        http://www.ucolick.org/~sla/     Hgt +250 m