Help with ISCSI SAN
Paul M. Dyer
pmdyer at ctgcentral2.com
Mon Feb 1 12:13:14 CST 2010
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 scms.waikato.ac.nz>
To: linux-poweredge at dell.com
Sent: Tuesday, January 12, 2010 9:09:09 PM (GMT-0600) America/Chicago
Subject: Help with ISCSI SAN
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 dell.com
Please read the FAQ at http://lists.us.dell.com/faq
More information about the Linux-PowerEdge