I’ve been chasing down a problem for a few days on a SLES 11 load. I’ve tried basic mdadm as well as the “Intel RAID”. Modified some of the mkinitrd scripts so that it doesn’t error out, and actually builds the initrd.
But it never includes the mdadm or the /etc/mdadm.conf files. So the boot with the new initrd can’t assemble the raid correctly, and can’t do a correct switchroot to the raid device.
This is annoying.
I pointed out here before that sometimes what distros do to differentiate themselves eventually winds up making things really bad. Be this poor choices of utilities and tools, or bad configuration options. Ubuntu has suffered from this disease as of late. 9.10 isn’t good.
SLES11 has its upsides. Some of it is good. But not allowing software RAID1 on OS drives?
Ok, in all fairness, RHEL 5 doesn’t do this right for non-RH kernels … they do some special magic with dmraid. I don’t want dmraid, I want mdraid.
So we patched the mkinitrd script on RHEL. Patch works great, forces mdadm and /etc/mdadm.conf into the initrd. RAIDs assemble correctly, and things work.
Well, we’ve been edging towards a diskless config on these units (either iSCSI, NFS, or RAMDISK based). Its looking more and more like this is the right path over time. Stateful installations are requiring traversal of installation stacks that break important functionality. Why use them if they are going to break.
So off I go to try to work around this problem.