Help with ISCSI SAN

Paul M. Dyer pmdyer at
Mon Feb 1 12:13:14 CST 2010

Hi Clint,

My feeling is that the NIC on the server is not where most multipath happens.   You could use bonding for 2 NICs to have a secondary NIC take over if the path from server NIC-1 gets disconnected.

On the other end is where multipath takes place.   On a setup that I have, multipath logs in to 4 different paths on the SAN, and shows the paths with "multipath -l":

xr (36006016025302300a2f50e7c0981de11) dm-10 DGC,RAID 1
[size=24G][features=1 queue_if_no_path][hwhandler=1 emc][rw]
\_ round-robin 0 [prio=0][active]
 \_ 1:0:0:6 sdz  65:144 [active][undef]
 \_ 3:0:0:6 sdaa 65:160 [active][undef]
\_ round-robin 0 [prio=0][enabled]
 \_ 2:0:0:6 sdab 65:176 [active][undef]
 \_ 4:0:0:6 sdac 65:192 [active][undef]


----- Original Message -----
From: "Clint Dilks" <clintd at>
To: linux-poweredge at
Sent: Tuesday, January 12, 2010 9:09:09 PM (GMT-0600) America/Chicago
Subject: Help with ISCSI  SAN 

Hello Everyone

I am wanting feed back from people familiar with Dell Equallogic ISCSI SANS.

Our organization has just purchased a  Dell Equallogic ISCSI SAN with 
all the equipment required to allow a host to have fully redundant paths 
to this SAN.

We have also purchased PowerEdge Servers with Broadcom NetXtremeII Nics 
/ISCSI HBA's.  On some of these servers we want to run Redhat Enterprise 
Linux 5.4 or CentOS 5.4

I can find documentation on using Software ISCSI initiators, but nothing 
on using Broadcom Hardware HBA's so I am assuming it is a case of go 
into the BIOS of the server put in the license key for the Hardware HBA 
and then another device that looks like a Nic will be present  once RHEL 
or CentOS is installed.  Is this correct ?

If this is the case I am assuming I still use the open iscsi initiator 
tools to manage things ?

To ensure redundancy each Linux host will have two nics going connected 
to the ISCSI network. And I am aware that I can use dmsetup / multipath 
to setup a redundant device but one thing that confuses me is that if 
the two nics (each connected to a separate switch) are on the same 
subnet (as Dell have instructed me to  do)  how can you guarantee that 
the two connections you see are using the appropriate redundant paths ?  
Or is this not really an issue ?

Thank you for any thoughts or experiences that you can share related to 
a setup like this, and have a nice day :)

Linux-PowerEdge mailing list
Linux-PowerEdge at
Please read the FAQ at

More information about the Linux-PowerEdge mailing list