Page MenuHomePhabricator

mw2202/mw2203 failed to install
Closed, ResolvedPublic

Description

mw2202 and mw2203 failed to install, after a reboot it cannot access the disks:

Loading Linux 4.4.0-1-amd64 ...
Loading initial ramdisk ...
Loading, please wait...
mdadm: No devices listed in conf file were found.
Gave up waiting for root device. Common problems:

  • Boot args (cat /proc/cmdline)
    • Check rootdelay= (did the system wait long enough?)
    • Check root= (did the system wait for the right device?)
  • Missing modules (cat /proc/modules; ls /dev)

ALERT! /dev/disk/by-uuid/84100d13-cfff-4ebd-b5bb-f8f402ddf5f7 does not exist. Dropping to a shell!
modprobe: module ehci-orion not found in modules.dep

BusyBox v1.22.1 (Debian 1:1.22.0-9+deb8u1) built-in shell (ash)
Enter 'help' for a list of built-in commands.

/bin/sh: can't access tty; job control turned off
(initramfs)

Could you run a hardware check on both systems for broken disks/controller?

mw2200/mw2201 from the same set of hardware class worked fine and I also tried booting with an older kernel, so probably a hardware issue.

Event Timeline

Restricted Application added subscribers: Southparkfan, Aklapper. · View Herald TranscriptSep 7 2016, 11:37 AM
elukey added a subscriber: elukey.Sep 7 2016, 4:28 PM

Had the same problem with 220[567] and fixed it adding manually rootdelay=60 at boot.

@MoritzMuehlenhoff Do we stay have to work on this?

MoritzMuehlenhoff closed this task as Resolved.Sep 14 2016, 2:42 PM

The installation of these and other servers failed since the kernel doesn't detect the disks within the default time span. This also happened on mw2205-mw2210, mw2212 and mw2139. There's a workaround (setting rootdelay manually in the grub config). For some reason the disk setup takes longer on some of the servers, but it's pretty random on otherwise identical hardware. Since the subset of affected servers is manageably low, let's just ignore this. We change d-i to use a higher delay by default,but it's probably not worth the efforts. I'll close this bug since a workaround exists.