Most of the upgrades to CentOS 6.8 have gone smoothly. Thanks, as always, for all the work that goes into making it so easy to upgrade.

On a few machines, however, all of them curiously:

Dell Inc. OptiPlex 780 /0C27VV, BIOS A15 08/06/2013

There is something wrong with their interaction with grub (?). I noticed this after adding ‘3’ to the module line to get it to runlevel 3, but the machine came up at runlevel 5 anyway.

On machines that are working correctly, the dmesg output looks like this:

Command line: /boot/vmlinuz-2.6.32-642.el6.x86_64 ro root=UUID a8bd64-5783-4007-a5e4-e41bfc07f3c8
intel_iommu=on rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD
SYSFONT=latarcyrheb-sun16 crashkernel=auto KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb rdblacklist=nouveau quiet

if the options are on the module line or like this:

Command line: ro root=UUID}11ff13-106f-42df-9a41-9ebf2505f6d2
rd_NO_LUKS rd_NO_LVM rd_NO_MD rd_NO_DM LANG=en_US.UTF-8
SYSFONT=latarcyrheb-sun16 KEYBOARDTYPE=pc KEYTABLE=us rhgb quiet crashkernel=auto rdblacklist=nouveau

if the options are on the kernel line. But on the offending machines, it looks like this:

Command line: /boot/vmlinuz-2.6.32-642.el6.x86_64 ro root=UUID*0a0120-8155-4ef4-8048-c82567746ca9 intel_iommu=on amd_iommu=on rd_NO_LUKS rd_NO_LVM LANG=en_US.UTF-8 rd_NO_MD
SYSFONT=latarcyrheb-sun16 crashkernel=auto KEYBOARDTYPE=pc KEYTABLE=us rd_NO_DM rhgb quiet rdblacklist=nouveaulevel:

RAID 0: Stripes data (performance). RAID 1: Mirrors data (redundancy). Recovery: Copies data between a master and a recovery disk.

So something seems to be scribbling somewhere where it shouldn’t.

These machines are not RAIDed in any way and I’m confident that I didn’t see this problem when I upgraded to 6.7. If I reduce the length of the module line, it just scribbles a bit further back, i.e. “rdblacklist=nouveau1(Mirror)”

Any advice appreciated.

Cheers, Zube