ISCSI (swiscsi) clarification in Esxi 4.0.

[TraversiRoberto]

Hello,

i'd like to ask some clarification about the configuration procedure described in http://www.vmware.com/pdf/vsphere4/r40/vsp_40_iscsi_san_cfg.pdf (page 36 example 2-1).

I've set up two Freenas box with one ISCSI target each, at the moment i'm pointing to them only from one Esxi server on four different subnets (below the details):

Esxi:

vmk2     1.1.1.x 255.255.255.0     (vswitch0, overriding default active active teaming and setting vmnic0 active and vmnic1 standby (not none as suggested by documentation))

vmk3     2.2.2.x 255.255.255.0     (vswitch0, overriding default active active teaming and setting vmnic0 standby and vmnic1 active (not none as suggested by documentation))

vmk4     3.3.3.x 255.255.255.0     (vswitch1, overriding default active active teaming and setting vmnic2 active and vmnic3 standby (not none as suggested by documentation))

vmk5     4.4.4.x 255.255.255.0     (vswitch1, overriding default active active teaming and setting vmnic2 standby and vmnic3 active (not none as suggested by documentation))

vmnic0 bandwidth 5000Mbps (HP blade with virtualconnect)

vmnic1 bandwidth 5000Mbps (HP blade with virtualconnect)

vmnic2 bandwidth 4000Mbps (HP blade with virtualconnect)

vmnic3 bandwidth 4000Mbps (HP blade with virtualconnect)

FN1:

em0     1.1.1.y 255.255.255.0

em1     2.2.2.y 255.255.255.0

em2     3.3.3.y 255.255.255.0

em3     4.4.4.y 255.255.255.0

em0 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em1 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em2 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em3 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

FN1:

em0     1.1.1.z 255.255.255.0

em1     2.2.2.z 255.255.255.0

em2     3.3.3.z 255.255.255.0

em3     4.4.4.z 255.255.255.0

em0 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em1 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em2 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

em3 bandwidth 1000Mbps (Intel pro1000 MF pci-e)

The question is:

at the moment everything is working fine including mpio (vmware round robin), having vmnic0/1 in active standby as well as vmnic2/3 i can continue using mpio on all four subnet even if an uplink of the blade virtualconnect blade goes down, if i follow documentation putting the standby vmnic to none state i'll lose that situation (obviously mpio permits however to make traffic on 3 Gbps link) .... but what i gain with the procedure described as "activating multipathing for software iscsi initiator" which suggests the following:

esxcli swiscsi nic add -n vmk2 -d vmhba32
esxcli swiscsi nic add -n vmk3 -d vmhba32
esxcli swiscsi nic add -n vmk4 -d vmhba32
esxcli swiscsi nic add -n vmk5 -d vmhba32

Thank you in advance for your help.

Best regards, R.T.

Messaggio modificato da TraversiRober…