booting from MD Device leaves raid degraded

This document (7018240) is provided subject to the disclaimer at the end of this document.

Environment

SUSE Linux Enterprise Server 11

Situation

A SLES 11 System was installed with the root Filesystem on a MD Raid Device. Sometimes after a boot the there is a path missing from the MD, so the Raid is degraded.

Manually re-adding the device works:

# mdadm --manage /dev/md0 -a /dev/myshinyonep2
mdadm: re-added /dev/myshinyonep2

Resolution

add the lines

AUTO -all
POLICY path=pci-0000:YY:* action=re-add
POLICY path=pci-0000:YY:* action=re-add

to /etc/mdadm.conf. Please keep in mind that the above

YY

has to be found out from the system. These values have to be adjusted
based on the PCI path the devices are actually on. If in doubt you can
run

supportconfig

and consult the resulting hw.txt

Run

mkinitrd

to update the md configuration in the initial ramdisk.

After this a reboot will verify that the issue is resolved.

Cause

The issue in this case is that the md will not wait for all devices to appear but will press
on with the booting process for speed. To build a dependency like it would be possible
in SLE 12 is not feasable in SLE 11. The solution in this case is to let the system boot
but to ensure that the MD Array will be reassembled automatically.

Disclaimer

This Support Knowledgebase provides a valuable tool for SUSE customers and parties interested in our products and solutions to acquire information, ideas and learn from one another. Materials are provided for informational, personal or non-commercial use within your organization and are presented "AS IS" WITHOUT WARRANTY OF ANY KIND.