multipath config problems?

[Richard Amick]

I have the following hardware:
2 node redundant cluster with shared disks consisting of:
2x ProLiant DL380R03 servers each with:
2x 2.8GHz Xeon 512K cache processors
6G memory
6x 18.2GB pluggable 1" Ultra3 SCSI 15K drives

Smart Array 500 Redundant Disk Controller:
4 Port Shared Storage Module w/Smart Array Multipath Software
2 Smart Array Controllers each with 512M battery backed cache
14x 36.4GB Pluggable 1" Ultra3 SCSI 15K drives
NC7770 PCI-X Gigabit Ethernet interconnect and external connect (parallel interconnect for redundancy)

and the following software:
O/S - RedHat Linux AS 2.1 (2.4.9-e.27)

I am trying to get multipath set up correctly on these servers.
The following is (part) of my /etc/raidtab:
raiddev /dev/md0
raid-level multipath
nr-raid-disks 2
persistent-superblock 1
chunk-size 8
device /dev/cciss/c0d3
raid-disk 0
device /dev/cciss/c1d0
raid-disk 1

The following is the output of lsmod:
Module Size Used by Not tainted
ocfs 306304 15
bcm5700 116996 3
bonding 27116 2
multipath 11940 7 (autoclean)
ext3 74176 7
jbd 55304 7 [ext3]
cciss 64032 17
sd_mod 13888 0 (unused)
scsi_mod 126412 1 [cciss sd_mod]

This is what I'm getting at startup (revelant dmesg):
md: autorun ...
md: considering cciss/c0d3 ...
md: adding cciss/c0d3 ...
md: created md0
md: running:
RAID level -4 does not need chunksize! Continuing anyway.
md: multipath personality registered as nr 7
md0: max total readahead window set to 124k
md0: 1 data-disks, max readahead per data-disk: 124k
multipath: device cciss/c0d3 operational as IO path 0
(checking disk 0)
multipath: array md0 active with 1 out of 1 IO paths (0 spare IO paths)
md: updating md0 RAID superblock on device
md: ... autorun DONE.

So, shouldn't it be considering /dev/cciss/c1d0 and have:
multipath: array md0 active with 1 out of 2 IO paths (1 spare IO paths)

[Dineshkumar Surpur]

I am also running into the same problem, Richard were you able to get any solution let me know too.

The problem is i have 4 paths connected to a exported lun say for ex seen as sdf, sdk, sdp and sdu are the same storage lun

i did the devlabel

# fdisk and used the t option to specify it as a raid disk for all devices sdf,sdk,sdu and sdp
# devlabel add -d /dev/sdf1 -d /dev/sdk1 -d /dev/sdu1 -d /dev/sdp1 -s /dev/develop --multipath --automount
# edited /etc/raidtab
raiddev /dev/md0
raid-level multipath
nr-raid-disks 1
chunk-size 32
device /dev/develop_multipath0
raid-disk 0

# mkraid /dev/md0
# raidstart


#cat /proc/mdstat
Personalities : [multipath]
read_ahead 1024 sectors
Event: 1
md0 : active multipath sdf1[0]
4192832 blocks [1/1] [U]

shows me only one device and when i do I/O to /dev/md0 the I/O goes only on one path and not all the paths ? How do i configure the /etc/raidtab using multipath option so that I/O is done using multiple paths for a given disk ?