Start a Conversation

Unsolved

This post is more than 5 years old

1583

September 25th, 2011 03:00

EMC Powerpath failover not working

HI,

have an "Boot from SAN" environment (running on IBM Bladecenter with EMC CX3 SAN), where failover isn't working.

Each Blade has partitions from two different EMC's provided. One partition has hosted the OS,

second one is a "data" partition.

OS is SLES10 SP3 x84_64 and EMC Powerpath Version: EMCPower.LINUX-5.5.0.00.00-275.SLES10SP3.x86_64.rpm

Reboot  showed GRUB 22 error straight away "Must load Multiboot kernel before modules"

Konfiguration seems to be ok for single SAN environment, same configuration is working well, when only one

EMC (boot/data combined on one partition) is connected.

Any comment or hint to identify the root cause is much appreciated !

Thansk, best regards

Christoph

Infos/outputs:

# powermt display dev=all

Pseudo name=emcpowera

CLARiiON ID=APM00091102568 [node1]

Logical device ID=6006016019E1220033ACD17F15DEDF11 [LUN 280]

state=alive; policy=CLAROpt; priority=0; queued-IOs=0;

Owner: default=SP A, current=SP A       Array failover mode: 1

==============================================================================

--------------- Host ---------------   - Stor -   -- I/O Path --  -- Stats ---

###  HW Path               I/O Paths    Interf.   Mode    State   Q-IOs Errors

==============================================================================

  1 qla2xxx                 sdd       SP A0     active  dead        0      1

   2 qla2xxx                  sdf       SP B1     active  alive       0      0

   2 qla2xxx                  sdg       SP A1     active  alive       0      0

  1 qla2xxx                  sdh      SP B0     active  dead        0      1

Reboot showed GRUB 22 error straight away "Must load Multiboot kernel before modules"--

/etc/sysconfig/kernel output:

## Path:        System/Kernel

## Description:

## Type:        string

## Command:     /sbin/mkinitrd

#

# This variable contains the list of modules to be added to the initial

# ramdisk by calling the script "mkinitrd"

# (like drivers for scsi-controllers, for lvm or reiserfs)

#

## Type:        string

INITRD_MODULES="mptsas qla2xxx processor thermal fan jbd ext3 dm_mod edd"

node1: # uname -r

2.6.16.60-0.77.1-smp

2 Intern

 • 

1.3K Posts

September 25th, 2011 06:00

do you know why A0 and B0 path dead? In my memory, the BOOT luns are typically hard coded at BIOS level to boot from a particular path/SP.

13 Posts

September 25th, 2011 20:00

We're testing failover, therefore we disabled the first McData SAN switch, and performed an reboot.

That reboot is failing.

62 Posts

September 29th, 2011 23:00

what is the Failover mode for this hosts and what flare version on the Clariion??

13 Posts

September 30th, 2011 05:00

failover : policy=CLAROpt; Flare 3.26.80.5.011

62 Posts

October 1st, 2011 04:00

I am talking abt the Failover mode.Go to the connectivity status and check for details on the respective hosts...

No Events found!

Top