[gridengine users] Leap second and SGE 6.2u5

[Shawn Stephens]

That fixed it!

date -s "`date`"

Thanks Kasper!

Here's the patch info for the kernel: http://marc.info/?l=linux-kernel&m=134113577921904&w=2

Shawn

On Jul 1, 2012, at 11:15 AM, Kasper Fischer wrote:

Dear Laurence, dear Shawn,

I have read that many people had this problem with mysql servers. If I understand the related blogs right it is a kernel problem. The solution was just to reset the clock of the machine, e. g. running date -s "`date`" (or ntpdate). I hope this helps and you can avoid rebooting.

BTW: I didn't observe any increased loads on my hosts (running Debian).

Regards,

Kasper
 +----------------------------------------------------------------------+
| Dr. Kasper D. Fischer            <kasper.fischer at ruhr-uni-bochum.de> |
| Ruhr-Universität Bochum                                              |
| Institut für Geologie, Mineralogie und Geophysik                     |
| NA 3 / 174                       Tel: +49 234 3227574                |
| D-44780 Bochum                   Fax: +49 234 3214181                |
| Germany                          www.geophysik.ruhr-uni-bochum.de    |
+----------------------------------------------------------------------+
Am 01.07.2012 16:43, schrieb Laurence Mayer:
> I am having very similar problems.
> OS: Ubuntu 10.04
> SGE: sge6_2u3
> 
> CPU utilization of sge_execd is between 44% and 175% which is much much higher than usual, started with the leap.
> 
> I have stopped and started sge_execd with no success and I was required to reboot the nodes to reduce the CPU utilization.
> 
> I have still a couple of nodes that require reboot, but due to jobs on these nodes I am unable to do so.
> 
> Would prefer to find a solution rather than reboot.
> 
> Thanks
> Laurence
> 
> On Sun, Jul 1, 2012 at 4:27 PM, Shawn Stephens <shawn.stephens at gmail.com> wrote:
> I noticed last night that all of my ge daemons ramped up CPU consumption right after the leap second was inserted.  My jobs seem to be working fine, but there's a little more system cpu load from sge_execd.  sge_qmaster is running at around 200% CPU, but everything seems to be functioning.  I'm trying to figure out a way to restart all of the ge daemons without impact to my running and queued jobs.  Anybody else see this problem?
> 
> compute nodes:
> OS: CentOS 6.2
> SGE: gridengine-execd-6.2u5-6.el6.x86_64
> 
> qmaster:
> OS: Fedora 15
> SGE: gridengine-qmaster-6.2u5-10.fc15.3.x86_64
> 
> Shawn
> 
>                    ^
>                  /   \
>                /       \
>               ^         ^
>             /   \     /   \
>            |     |   |     |
>            ^     ^   ^     ^
>           / \   / \ / \   / \
>           | |   | | | |   | |
> =========================================
>   ___         ___  
>  / __|___ ___| __|_ _  ___ _ _ __ _ _  _ 
> | (_ / -_) _ \ _|| ' \/ -_) '_/ _` | || |
>  \___\___\___/___|_||_\___|_| \__, |\_, |
>                               |___/ |__/  
> Shawn Stephens
> Senior IT Administrator
> GeoEnergy, Inc.
> 3100 Wilcrest Dr., Suite 220
> Houston, TX 77042 USA
> +1 (713) 539-3955 mobile
> 
> _______________________________________________
> users mailing list
> users at gridengine.org
> https://gridengine.org/mailman/listinfo/users
> 
> 
> 
> 
> _______________________________________________
> users mailing list
> users at gridengine.org
> https://gridengine.org/mailman/listinfo/users


<kasper_fischer.vcf>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://gridengine.org/pipermail/users/attachments/20120701/4f823f00/attachment-0001.html>