[emanicslab] EmanicsLab - Status Update

[Rick Hofstede]

Hi David,

Do you have any news on the CPU utilization problem of emanicslab1.ewi.utwente.nl? As of today (or the end of last week) we also have problems with emanicslab2.ewi.utwente.nl; people effectively cannot login anymore. One of the users receives the following error message:

Last login: Thu Apr 17 09:15:58 2014 from wlan021251.mobiel.utwente.nl
WARNING - setns(/proc/627/ns/uts)=>-2 (ignored)
WARNING - setns(/proc/627/ns/ipc)=>-2 (ignored)
WARNING - setns(/proc/627/ns/net)=>-2 (ignored)
WARNING - setns(/proc/627/ns/mnt)=>-2 (ignored)
Traceback (most recent call last):
  File "/usr/sbin/lxcsu-internal", line 273, in <module>
    main()
  File "/usr/sbin/lxcsu-internal", line 220, in main
    if (not umount(fs_path,'-l')):
  File "/usr/sbin/lxcsu-internal", line 42, in umount
    output = os.popen('/bin/umount %s %s 2>&1'%(opts, fs_dir)).read()
OSError: [Errno 12] Cannot allocate memory
Connection to emanicslab2.ewi.utwente.nl closed.

Best regards,

--
Rick

On 09 Apr 2014, at 13:26, David Hausheer <hausheer at ps.tu-darmstadt.de> wrote:

> Dear Andri,
> 
> Thanks for the note. I am aware of the CPU utilization problem, which seems to originate from systemd. I am looking into that and I hope to be able to solve it as soon as possible.
> 
> Best regards
> David
> 
> On 09.04.2014 09:42, Andri Lareida wrote:
>> Dear David,
>> 
>> There seems to be a problem with the first node in Twente:
>> emanicslab1.ewi.utwente.nl
>> After the upgrade I was not able to copy files with SCP to the node, I
>> was able to log in though. Guilherme re-created my slice and now I can
>> not log in anymore. Guilherme also reported that there is a very high
>> CPU load on the machine and he can see processes from all users.
>> Something seems to be wrong with the set up.
>> Thanks for investigating.
>> 
>> Cheers
>> Andri
>> 
>> On 04/08/2014 01:50 AM, David Hausheer wrote:
>>> Dear all,
>>> 
>>> here is another status update of migration of EmanicsLab nodes:
>>> 
>>> UZH - both nodes running and up-to-date
>>> TUD - both nodes running and up-to-date
>>> PSNC - both nodes running and up-to-date
>>> UCL - both nodes running and up-to-date
>>> IBBT - both nodes running and up-to-date
>>> INRIA - both nodes running and up-to-date
>>> UPC - both nodes running and up-to-date
>>> UniGE - both nodes running and up-to-date
>>> UT - both nodes running and up-to-date
>>> JUB - both nodes running and up-to-date
>>> UniBW - both nodes running and up-to-date
>>> 
>>> We still need to install Ganglia on the nodes.
>>> 
>>> Also, we are currently facing a problem with CPU utilization (some of
>>> the nodes run at 100% CPU). We hope to be able to fix it soon.
>>> 
>>> Best regards
>>> David
>>> 
>> 
>> 
>> 
>> _______________________________________________
>> emanicslab mailing list
>> emanicslab at lists.ifi.uzh.ch
>> https://lists.ifi.uzh.ch/listinfo/emanicslab
>> 
> 
> -- 
> Prof. Dr. David Hausheer
> 
> Technische Universitaet Darmstadt
> Dept. of Electrical Engineering & Information Technology
> 
> Rundeturmstr. 10, Building S3/20, Room 225
> 64283 Darmstadt, Germany
> Phone: +49 6151 16 4280
> Fax: +49 6151 16 6152
> E-Mail: hausheer at ps.tu-darmstadt.de
> Web: http://www.ps.tu-darmstadt.de/
> _______________________________________________
> emanicslab mailing list
> emanicslab at lists.ifi.uzh.ch
> https://lists.ifi.uzh.ch/listinfo/emanicslab

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ifi.uzh.ch/pipermail/emanicslab/attachments/20140422/b893c81a/attachment.html>