Hello everybody,

We have a server with a dual Path HBA. We have two fabrics and the server can access two LUNs using 4 paths (500 Gb and 10 Gb LUNs). We are sure about our SAN configuration. This is running on a Debian system using Kernel 3.0.8. This is the dmesg output:


Code:
[ 157.534566] sd 1:0:1:0: [sde] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 157.534569] sd 1:0:1:0: [sde] Sense Key : Illegal Request [current]
[ 157.534572] sd 1:0:1:0: [sde] <<vendor>> ASC=0x94 ASCQ=0x1ASC=0x94 ASCQ=0x1
[ 157.534576] sd 1:0:1:0: [sde] CDB: Read(10): 28 00 00 00 10 00 00 00 08 00
[ 157.534585] end_request: I/O error, dev sde, sector 4096
[ 157.534591] sd 2:0:0:1: [sdh] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 157.534593] sd 2:0:0:1: [sdh] Sense Key : Illegal Request [current]
[ 157.534596] sd 2:0:0:1: [sdh] <<vendor>> ASC=0x94 ASCQ=0x1ASC=0x94 ASCQ=0x1
[ 157.534600] sd 2:0:0:1: [sdh] CDB: Read(10): 28 00 00 00 10 00 00 00 08 00
[ 157.534606] end_request: I/O error, dev sdh, sector 4096
[ 157.534612] sd 2:0:0:0: [sdg] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
[ 157.534615] sd 2:0:0:0: [sdg] Sense Key : Illegal Request [current]
[ 157.534618] sd 2:0:0:0: [sdg] <<vendor>> ASC=0x94 ASCQ=0x1ASC=0x94 ASCQ=0x1
[ 157.534622] sd 2:0:0:0: [sdg] CDB: Read(10): 28 00 00 00 10 00 00 00 08 00
[ 157.534628] end_request: I/O error, dev sdg, sector 4096
Is this a normal startup secuence?. Are those errors normal messages?

Those are multipath outputs:


Code:
3600a0b80006e0afa0000f9ad50c5ac85 1:0:0:0 sdc 8:32 6 undef ready IBM,1818
3600a0b80006e0b0a0000007f511879ac 1:0:0:1 sdd 8:48 6 undef ready IBM,1818
3600a0b80006e0afa0000f9ad50c5ac85 1:0:1:0 sde 8:64 1 undef ghost IBM,1818
3600a0b80006e0b0a0000007f511879ac 1:0:1:1 sdf 8:80 1 undef ghost IBM,1818
3600a0b80006e0afa0000f9ad50c5ac85 2:0:0:0 sdg 8:96 1 undef ghost IBM,1818
3600a0b80006e0b0a0000007f511879ac 2:0:0:1 sdh 8:112 1 undef ghost IBM,1818
3600a0b80006e0afa0000f9ad50c5ac85 2:0:1:0 sdi 8:128 6 undef ready IBM,1818
3600a0b80006e0b0a0000007f511879ac 2:0:1:1 sdj 8:144 6 undef ready IBM,1818
Code:
3600a0b80006e0afa0000f9ad50c5ac85 dm-0 IBM,1818 FAStT
size=500G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
|-+- policy='round-robin 0' prio=6 status=active
| |- 1:0:0:0 sdc 8:32 active ready running
| `- 2:0:1:0 sdi 8:128 active ready running
`-+- policy='round-robin 0' prio=1 status=enabled
|- 1:0:1:0 sde 8:64 active ghost running
`- 2:0:0:0 sdg 8:96 active ghost running
3600a0b80006e0b0a0000007f511879ac dm-1 IBM,1818 FAStT
size=10G features='1 queue_if_no_path' hwhandler='1 rdac' wp=rw
|-+- policy='round-robin 0' prio=6 status=active
| |- 1:0:0:1 sdd 8:48 active ready running
| `- 2:0:1:1 sdj 8:144 active ready running
`-+- policy='round-robin 0' prio=1 status=enabled
|- 1:0:1:1 sdf 8:80 active ghost running
`- 2:0:0:1 sdh 8:112 active ghost running
We have tested to mount /dev/mapper/3600a0b80006e0b0a0000007f511879ac. Then we start a 9,2 Gb file copy operation. During the copy process, we physically remove a fibre channel cable (port1), wait for 10 seconds, insert the fibre channel cable, wait 10 seconds, remove the other fiber channel cable (port2), wait 10 second and insert again the cable. The copy process was succesfully ended.

We think the multipath configuration is working the rigth way. Are we wrong?

Thanks in advance. (Sorry about my english).