Leapsecond

username_taken · 4th July 2012, 1:32 PM

Anyone suffer any damage from the Leapsecond on saturday night ?

we had a few appliances that were obviously running older linux kernels in the background kick up their CPUs and need rebooting ... but otherwise not too bad.

digamma · 4th July 2012, 1:57 PM

Seems someone did have a few issues, also with linux based os's.

http://www.cio.com.au/article/429343.../?fp=4&fpid=16

username_taken · 5th July 2012, 4:17 AM

anyone running a rhel 6 variant is possibly affected. bigger servers hide it better for up to a few days before kernel crashing out. if you're running rhel6 variants I would suggest checking syslog for messages about cpu softlocks, and checking process run times for anything that reckons its been running for thousands/millions of hours/days.

there's a good thread here talking about it

http://serverfault.com/questions/403...-a-leap-second

plasticbastard · 5th July 2012, 9:40 AM

We've had the issue affect our SLES servers. It made a noticeable dent on one server in particular (it has serves a fair bit of java based stuff), with average load at 39. Applying a new date to the machine fixed the issue straight away, with the load returning back to less than 0 within a minute.

No reboot for us was necessary until we came across the issue again on the same server later the next day. So we've restarted the server & all seems much better.

PabloEscobar · 5th July 2012, 10:00 AM

If this happened to Windows Servers, the Linux fanbois would be all over it.

.

The Qantas outage made the news, But AFAIK, the Telstra "Cloud" hosting services also had issues on the weekend, But there is no mention of it anywhere.

mike-s · 15th July 2012, 10:21 PM

I'm glad I left my old work before this went down. They run a lot of the Qantas servers and I can imagine the chaos this would have caused.

4th July 2012, 1:32 PM	#1
username_taken Member Join Date: Oct 2004 Location: Austin, TX Posts: 1,293	Leapsecond Anyone suffer any damage from the Leapsecond on saturday night ? we had a few appliances that were obviously running older linux kernels in the background kick up their CPUs and need rebooting ... but otherwise not too bad.

4th July 2012, 1:57 PM	#2
digamma Member Join Date: Mar 2002 Location: Brisbane, Southside. Posts: 2,428	Seems someone did have a few issues, also with linux based os's. http://www.cio.com.au/article/429343.../?fp=4&fpid=16 __________________ The ultimate weakness of violence is that it is a descending spiral. Returning violence with violence only multiplies violence, adding deeper darkness to a night already devoid of stars.

15th July 2012, 10:21 PM	#6
mike-s Member Join Date: Dec 2003 Location: Sydney, Australia Posts: 1,696	I'm glad I left my old work before this went down. They run a lot of the Qantas servers and I can imagine the chaos this would have caused. __________________ Successfull trades: RakOon, drfbro, mR_CaESaR, Spyfox If I've got crap for sale on ebay you like, click here. Old hardware giveaway thread here. Sony Vaio Club Member #21

5th July 2012, 4:17 AM	#3
username_taken Member Join Date: Oct 2004 Location: Austin, TX Posts: 1,293	anyone running a rhel 6 variant is possibly affected. bigger servers hide it better for up to a few days before kernel crashing out. if you're running rhel6 variants I would suggest checking syslog for messages about cpu softlocks, and checking process run times for anything that reckons its been running for thousands/millions of hours/days. there's a good thread here talking about it http://serverfault.com/questions/403...-a-leap-second

5th July 2012, 9:40 AM	#4
plasticbastard Member Join Date: Jul 2003 Location: Sector ZZ9 Plural Z Alpha Posts: 3,941	We've had the issue affect our SLES servers. It made a noticeable dent on one server in particular (it has serves a fair bit of java based stuff), with average load at 39. Applying a new date to the machine fixed the issue straight away, with the load returning back to less than 0 within a minute. No reboot for us was necessary until we came across the issue again on the same server later the next day. So we've restarted the server & all seems much better.

5th July 2012, 10:00 AM	#5
PabloEscobar Member Join Date: Jan 2008 Posts: 2,817	If this happened to Windows Servers, the Linux fanbois would be all over it. . The Qantas outage made the news, But AFAIK, the Telstra "Cloud" hosting services also had issues on the weekend, But there is no mention of it anywhere.