perfSONAR User Q&A and Other Discussion

[Andrew Lake <>]

Hi,

See http://psps.perfsonar.net/toolkit/FAQs.html#Q64 for an explanation and 
the steps to correct this. It all stems from a problem in the version of 
anaconda that ships with CentOS 6 that ignores the usual partition directives 
you'd place in a kickstart file. As a result it requires us to re-arrange 
things later in the install process using a method that can run into issues 
if the volume name changes. IIRC one thing that can cause the volume name to 
change is assigning a DNS name during the install process. 

Thanks,
Andy




On Aug 1, 2014, at 8:12 AM, Peter Gutierrez 
<>
 wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> I'm building a few new perfsonar boxes, and grabbed
> http://software.internet2.edu/pS-Performance_Toolkit/pS-Performance_Toolkit-3.3.2-NetInstall-x86_64.iso
> 
> After installation, I found that the drive was partitioned with a root and
> a home partition:
> 
> $ df -h
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/mapper/vg_ps4-lv_root
>                       50G  1.6G   46G   4% /
> tmpfs                 7.8G     0  7.8G   0% /dev/shm
> /dev/sda1             485M   39M  421M   9% /boot
> /dev/mapper/vg_ps4-lv_home
>                      211G  188M  200G   1% /home
> 
> Back on January 8, 2014, on perfsonar-ps-users list, in a thread with the
> Subject: Re: [pS-PS-Users] Troubleshooting documentation, Andy replied
> about this happening then:
> 
> 
>> Ah ok. I think you may have ran into that problem if you installed one
>> of the early release candidates of 3.3 and then did a "yum update" to
>> the final release. One of the early release candidates gave all the
>> drive space over 20GB to /home by default. We caught this early on and
>> forced the NetInstall to give the space to root since that is where 
>> most of the data lives. Unfortunately I think the "fix" only applied to
>> new installs of 3.3 so if a release candidate tester went straight to
>> the final release then the drive layout was not altered.
> 
> 
> Any ideas on why this happened?
> 
> PeterG
> 
> - -- 
> Peter Gutierrez  OIT/Network Services  U. Massachusetts Amherst
> Voice: 413 545 1606   PGP: 0xE1668A41      
> 
> A: Because it interrupts the logical flow of discourse
> Q: Why is top posting bad?
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.11 (GNU/Linux)
> 
> iEYEARECAAYFAlPbhCoACgkQPTwtHOFmikF72QCff9r6kviSpna2ne0q3aJ+445F
> tyUAnRiPjRC1opd14AG2wOcG8FtyORxR
> =iYN8
> -----END PGP SIGNATURE-----