I am researching this issue.  Note that this was a "fastbug" that was 
released on July 19, 2016 .  So most will probably not install it.

--

Connie J. Sieh
Computing Services Specialist III

Fermi National Accelerator Laboratory
630 840 8531 office

http://www.fnal.gov
[log in to unmask]

On Thu, 28 Jul 2016, Steven Haigh wrote:

> That could probably be Connie atm.
>
> CC'ed into this thread.
>
> On 2016-07-28 16:08, Stijn De Weirdt wrote:
>> hi fredric,
>> =20
>> i confirm, things seem ok with the centos rpm
>> =20
>> quick inspection of the rpms also does not show something obviously
>> wrong (both rpms ship same files with same sizes).
>> =20
>> i took the centos rpms, modified the release using rpmrebuild
>> (rpmrebuild --release=3D18.el6_8.0.fromcentos -p thecentosrpms) for all
>> libcgroup rpms and added them to one of our local repos. yum is happy=20
>> again.
>> =20
>> how do we report this to the SL packagers? (or do we just assume they
>> read all these mails?)
>> =20
>> stijn
>> =20
>> On 07/27/2016 05:22 PM, SCHAER Frederic wrote:
>>> Same here.
>>> Hey ! This seems SL specific... !?
>>> =20
>>> I installed this one manually :=20
>>> CentOS/6.8/updates/x86_64/Packages/libcgroup-0.40.rc1-18.el6_8.x86_64.=
> rpm
>>> =20
>>> [root@dev7247 ~]# cat /etc/redhat-release
>>> Scientific Linux release 6.8 (Carbon)
>>> =20
>>> [root@dev7247 ~]# rpm -qi libcgroup
>>> Name        : libcgroup                    Relocations: (not=20
>>> relocatable)
>>> Version     : 0.40.rc1                          Vendor: CentOS
>>> Release     : 18.el6_8                      Build Date: Tue 12 Jul=20
>>> 2016 06:27:20 PM CEST
>>> Install Date: Wed 27 Jul 2016 05:15:28 PM CEST      Build Host:=20
>>> worker1.bsys.centos.org
>>> (...)
>>> =20
>>> And cgroups seem to still work whereas they failed with the SL RPM :
>>> [root@dev7247 ~]# service cgconfig restart
>>> Stopping cgconfig service:                                 [  OK  ]
>>> Starting cgconfig service:                                 [  OK  ]
>>> =20
>>> ?
>>> =20
>>> Regards
>>> =20
>>> -----Message d'origine-----
>>> De : [log in to unmask]
>>> [mailto:[log in to unmask]] De la part de=20
>>> ~Stack~
>>> Envoy=C3=A9 : mercredi 27 juillet 2016 12:59
>>> =C3=80 : Stijn De Weirdt <[log in to unmask]>;=20
>>> [log in to unmask]
>>> Objet : Re: sl6.8 libcgroup
>>> =20
>>> On 07/27/2016 03:53 AM, Stijn De Weirdt wrote:
>>>> hi all,
>>>> =20
>>>> =20
>>>> we have a update an sl67 node to sl68 (but not yet updated the=20
>>>> kernel),
>>>> and this updates
>>>> libcgroup-0.40.rc1-17.el6_7.x86_64
>>>> to
>>>> libcgroup-0.40.rc1-18.el6_8.x86_64
>>>> =20
>>>> however, it now seems that the cgonfigparser even fails to validate=20
>>>> the
>>>> distributed /etc/cgconfig.conf
>>>> =20
>>>>> [root@test2802 ~]# /sbin/cgconfigparser -l /etc/cgconfig.conf
>>>>> error at line number 17 at {:syntax error
>>>>> Error: failed to parse file /etc/cgconfig.conf
>>>>> /sbin/cgconfigparser; error loading /etc/cgconfig.conf: Have=20
>>>>> multiple paths for the same namespace
>>>> =20
>>>> =20
>>>> the /etc/cgconfig.conf is the same in both rpms
>>>> =20
>>>> anyone seeing this? or knows how to fix?
>>> =20
>>> Greetings,
>>> =20
>>> I discovered the exact same thing. I fully updated to 6.8 and rebooted
>>> into the new kernel. I haven't filed a bug report against it yet as I
>>> didn't have time yesterday to really dig into it. My "workaround" was=20
>>> to
>>> "yum downgrade libcgroup" on all my hosts until I could figure it out.
>>> =20
>>> ~Stack~
>>> =20
>>> =20
>
> --=20
> Steven Haigh
>
> Email: [log in to unmask]
> Web: https://www.crc.id.au
> Phone: (03) 9001 6090 - 0412 935 897
>