• Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 404
  • Last Modified:

multi storage protocol and vmkernel portgroup

VM and storage experts over there can give me an idea which way to design the switch?

Here is my situation, We have ver4.1 esx hosts which are connected to nfs and iscsi.
Both protocols are on same subnet and working ok for years (I do not know about ver4.1, but it is working eventhough they all in same network and link aggregation IP hash).
I am on board now to upgrade this network.
I was told to stay with same network layout as it is and upgrade the esx.
Now my problem is that  we have link aggregation with 3 physical nics.
I have iscsi and nfs on same subnet.

This is what I am thinking of doing
 create one vswitch with three vmkernel port as on the picture
for iscsi- one active and 2 unused-this will allow me to bind the vmkernel port
for NFS- three active

If i use one vmkernel port then I will put three nic active, but I can not bind vmkernelport to iscsi.
I will appreciate if you put your design,
Please note that I can layout the network with separate vlans.
  • 3
  • 2
1 Solution
When you add the additional physical adapters, you can configure the NIC Teaming settings granularly for the VMkernel ports.

Just go into the vSwitch properties, select port->Edit->NIC Teaming tab->Select "use explicit failover order"->select "Override vSwitch failover order."  Then you can set an active adapter and standby adapter(s).  Just set a different active/standby pair for each iSCSI port.
sara2000Author Commented:
Baes on th elink below: We can use nic teaming for iscsi, but
Do not use port binding ,
•If you use NFS and iSCSI storage on one host you should use NIC teaming for both and don`t use MPIO.

You shouldn't use any type of load balancing, round robin, or anything like that.  Once it establishes a path, VMware wants to continue using the same path.

It is a best practice to have multiple paths in case of a path failure, however.
sara2000Author Commented:
Is should select the nic teaming load balance policy to ip hash  should not I ?
The hashing load balancing policies only really work when you're communicating with multiple destinations.

If you're always talking to the same host, then both MAC hashing and IP hashing will always result in the same path (since it will always result in the same hash).

My personal preference for SAN communication is to just use Active/Standby failover.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

Join & Write a Comment

Featured Post

Train for your Pen Testing Engineer Certification

Enroll today in this bundle of courses to gain experience in the logistics of pen testing, Linux fundamentals, vulnerability assessments, detecting live systems, and more! This series, valued at $3,000, is free for Premium members, Team Accounts, and Qualified Experts.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now