Re: [HTCondor-users] centos 7 problem

[Dimitri Maziuk <dmaziuk@xxxxxxxxxxxxx>]

On 04/02/2015 11:40 AM, Tim Theisen wrote:
> Hi Dimitri,
> 
> I am not able to reproduce this problem.
...
> Can you shed any additional light on when might be going wrong?

Sorry Tim, I meant to check and post this but I got sidetracked:

[root@turkey ~]# mount -v
...
tmpfs on /run type tmpfs (rw,nosuid,nodev,mode=755)
...

As I suspected it's tmpfs (/var/run is a symlink to /run). Don't know
why you can't reproduce it -- this is a base centos 7.0 install with
nothing but condor (execute node only), I don't remember doing anything
special during installation & set up.

While we're at it, our condor:condor is in ldap and I've seen a query
timeout once or twice from
"ExecStartPre=-/usr/bin/chown -R condor:condor /var/run/condor"

I can put numeric ids in there but that'll be overwritten by next rpm
update. The only way out I can think of is copy condor:condor to
/etc/passwd|group. I guess I'll get to work on my ansible-fu.

-- 
Dimitri Maziuk
Programmer/sysadmin
BioMagResBank, UW-Madison -- http://www.bmrb.wisc.edu

Attachment: signature.asc
Description: OpenPGP digital signature