Bug 712360

Created attachment 504093 [details]
Storage log as of 06-08-2011

Created attachment 504094 [details]
Anaconda log on 06-08-2011

This is an EFI firmware consideration (we are quite familiar with DOC-17644 for
legacy PC BIOS multipath booting).

Our configuration is as follows:  
- IBM HS22v with UEFI firmware (not in legacy mode)
- 100% SAN storage, including boot with multipath
- Device layout is:  
-- /dev/mapper/mpathap1 => /boot/efi (VFAT)
-- /dev/mapper/mpathap2 => /boot (Ext4)
-- /dev/mapper/mpathap3 => VG (LVM2)
- /boot/grub/device.map has one line:  
-- (hd0) /dev/mapper/mpatha

The Anaconda installer in both RHEL 6.0 GA and 6.1 Beta refuses to install the
bootloader.  Returns the same error message (in console F5):  
  "could not open disk /dev/mpathap: no such file or directory"

The Anaconda rescue mode in RHEL 6.0 GA fails to create the /mnt/sysimage
mounts.  However, the Anaconda rescue mode in RHEL 6.1 GA does create the
/mnt/sysimage mounts proper (including /mnt/sysimage/boot and
/mnt/sysimage/boot/efi).

However, chroot into /mnt/sysimage and attempting to run grub-install with
various disk targets continually returns the common "no corresponding BIOS
drive".  It appears that GRUB 0.97-68 patched with EFI support does not know
how to break down a Multipath device into its raw EFI/GPT reference(s).

By default IBM SVC (2145) has ALUA turned on by default.

In /usr/share/doc/device-mapper-multipath-0.4.9/multipath.conf.defaults:  

#       device {
#               vendor                  "IBM"
#               product                 "2145"
#               getuid_callout          "/lib/udev/scsi_id --whitelisted
--device=/dev/%n"
#               features                "1 queue_if_no_path"
#               hardware_handler        "0"
#               path_selector           "round-robin 0"
#               path_grouping_policy    group_by_prio
#               failback                immediate
#               rr_weight               uniform
#               rr_min_io               1000
#               path_checker            tur
#               prio                    alua
#       }

E.g., on another system (using firmware legacy mode) to the same IBM SVC:  

# multipath -ll
mpathd (36005076801910158a000000000000069) dm-4 IBM,2145
size=30G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| |- 3:0:0:2 sdc 8:32  active ready running
| `- 4:0:0:2 sdk 8:160 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
  |- 3:0:1:2 sdg 8:96  active ready running
  `- 4:0:1:2 sdo 8:224 active ready running
mpathc (36005076801910158a00000000000006a) dm-1 IBM,2145
size=30G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| |- 3:0:1:3 sdh 8:112 active ready running
| `- 4:0:1:3 sdp 8:240 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
  |- 3:0:0:3 sdd 8:48  active ready running
  `- 4:0:0:3 sdl 8:176 active ready running
mpathb (36005076801910158a000000000000067) dm-7 IBM,2145
size=8.0G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| |- 3:0:0:1 sdb 8:16  active ready running
| `- 4:0:0:1 sdj 8:144 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
  |- 3:0:1:1 sdf 8:80  active ready running
  `- 4:0:1:1 sdn 8:208 active ready running
mpatha (36005076801910158a000000000000065) dm-0 IBM,2145
size=45G features='1 queue_if_no_path' hwhandler='0' wp=rw
|-+- policy='round-robin 0' prio=50 status=active
| |- 3:0:0:0 sda 8:0   active ready running
| `- 4:0:0:0 sdi 8:128 active ready running
`-+- policy='round-robin 0' prio=10 status=enabled
  |- 3:0:1:0 sde 8:64  active ready running
  `- 4:0:1:0 sdm 8:192 active ready running

Install RHEL6.1 on a multipathed storage device (IBM SVC) on an IBM HS22v
server.

Created attachment 504135 [details]
Screenshot of error

Hi,

this is the same problem as bug 696876, Anaconda does not accept 'mpatha' or 'mapper/mpatha' in the kickstart file.

Closing as a dup.

*** This bug has been marked as a duplicate of bug 696876 ***