VMware Cloud Community
tscott2340
Contributor
Contributor

Software iSCSI Initiator - Network Configuration - VMkernel Port Bindings. Active / Active vs Not Used / Last Active

So I have two hosts in the same cluster setup the same exact way.. One host is using both iSCSI vmKernel ports while the other isn't.. Can someone tell me what sets this?

Using software iSCSI to an EMC VNX 5300.

Thanks

ESX01     

  • Storage Adapters > Network Config > iSCSI_1 Not Used, iSCSI_2 Last Active
  • iSCSI_1 / vmk1 / 192.168.10.41 – Can ping all 8 SAN iSCSI IPs
  • iSCSI_2 / vmk1 / 192.168.10.141 - Can ping all 8 SAN iSCSI IPs
  • SAN can ping both host iSCSI IPs
  • NIC Teaming setup correctly. Dynamic and Static discovery setup the same as all others.

5-6-2014 12-09-52 PM.jpg

ESX02

  • Storage Adapters > Network Config > iSCSI_1 Active, iSCSI_2 Active
  • iSCSI_1 / vmk1 / 192.168.10.42 – Can ping all 8 SAN iSCSI IPs
  • iSCSI_2 / vmk1 / 192.168.10.142 - Can ping all 8 SAN iSCSI IPs
  • SAN can ping both host iSCSI IPs
  • NIC Teaming setup correctly. Dynamic and Static discovery setup the same as all others.

5-6-2014 12-12-16 PM.jpg

9 Replies
King_Robert
Hot Shot
Hot Shot

Configuring Software iSCSI Port Binding on ESXi 5

by admin

By default, the default iSCSI configuration creates only one path to each iSCSI target. Port binding is used to provide multiple paths to an iSCSI array. With vSphere 5, iSCSI port binding can be configured in the vSphere client as well as the CLI. Prior to ESXi 5, port binding could only be configured using the CLI. There are a few things to bear in mind before configuring port binding:

  • The array’s target ports must reside in the same layer 2 segment/IP subnet  as the vmkernel interfaces that will connect to it.
  • All VMkernel ports used for iSCSI must reside in the same layer 2 segment / IP subnet.
  • All VMkernel ports used for iSCSI must be on the same vSwitch.
  • Network routing is not supported. Enabling multipathing removes the ability to route to the iSCSI target.

iSCSI Network Configuration

The first step to configuring software iSCSI is to set up the appropriate network configuration:

  1. In the Configuration | Networking screen in the vSphere client, click Add Networking
  2. Select VMkernel and click Next
  3. Select Create a vSphere Standard Switch
  4. Select the adapters you wish to use for iSCSi traffic and click next. You will need at least two adapters.
  5. Label the Portgroup appropriately. Click Next.
  6. Configure an IP address for the VMkernel portgroup. This needs to be in the same IP subnet as the iSCSI targets, as discussed above. Click Next.
  7. Review the summary screen, then click Finish to create the new vSwitch.

The next step is to go into the settings of the new vSwitch and create additional portgroups for each nic you have assigned to the vSwitch. In my lab I used two nics so created two portgroups. My resulting configuration looked like this:

Reply
0 Kudos
doihavetoaddane
Contributor
Contributor

Could you  finish the instructions? there was no picture or link?

Reply
0 Kudos
cykVM
Expert
Expert

You could have googled for that Smiley Wink It's copied from the buildVirtual blog: Configuring Software iSCSI Port Binding on ESXi 5

Reply
0 Kudos
doihavetoaddane
Contributor
Contributor

Ah i have alreday see this. I though that you might have been showing something different. I  have 4 x 1GB connections coming from EMC CX3-40 ISCSi  in to ESXi 5.0, 5.1 and 5.5 hosts (experimenting here)  and i cannot get more that 2 x paths to show active. All documentation online shows 1 x path, 2 x path  but no 4 x path. Was possibly looking for something in ESXCLI to address volume sessions? I understand that there is a default for 2. Then not sure what to do about round robin.

10.1.2.18, 10.1.3.18, 10.1.4.18, 10.1.5.18,  all in same Vlan,

Reply
0 Kudos
julezatmortonbu
Enthusiast
Enthusiast

We are seeing an identical issue with 1 host in our Cisco UCS deployment.


Two ESXi 5.1 U3 hosts

Two iSCSI vmk NICs per host

Host 2 is fine.

Host 1 is the "bad" host but everything is configured similar to Host 2.

The only difference I'm seeing is the number of paths and connected targets is different.

Host 2 we're seeing 20 targets and paths (this seems odd since we only have 4 volumes, but it's still the one that's pathing correctly so I'm not going to ask questions)

Host 1 we're seeing 16 targets and paths.

I can vmkping both NICs on host 2 from host 1.
I can vmkping one NIC on host 1 from host 2 the other does not respond.
I can vmkping both NICs on host 1 from host 1.

Things I've tried and have not worked:
I have tried to remove the vmkNIC iSCSI binding on the bad vmkNIC and re-add the binding.
I have tried a different IP address on the bad vmkNIC.

I have actually created a new vmkNIC using the same vmnic as the old vmkNIC using the same and different IP.

No matter what I do I get the same results seen in the photos at the start and below.

Again only a single host in our UCS deployment has this issue, and it's using the same UCS template from what I can tell across the board.
VLAN is correct, etc...

Good host 2, Bad host 1.
2015-07-22_13-24-50.jpg2015-07-22_13-22-57.jpg


As you can see active/unused is setup correctly, this is of course flipped on the other vmk for iSCSI.

2015-07-22_13-26-25.jpg

UCS traffic on the vmnics used for iSCSI A and B legs.  Obvious one isn't being used.

2015-07-22_10-38-18.jpg

Reply
0 Kudos
akarydas2
Enthusiast
Enthusiast

Do you have configured the Round Robin at the datastores paths?

pastedImage_0.png

Regards

Reply
0 Kudos
julezatmortonbu
Enthusiast
Enthusiast

No round robin, our array has it's own pathing driver which looks correct.  I think we've actually whittled it down to either A. a problem with the iSCSI initiator in VMware or B. our UCS config since the only hosts seeing any issues are UCS.  All non-UCS hosts seem to have no issues.
Unfortunately we can't reboot these for awhile since the hosts having the issue are Cisco UC for the company which will require some scheduled down time.  We have Cisco going over our UCS config to make sure there's nothing awry there first, and if not then we're going to reboot the hosts.

Reply
0 Kudos
dprows33
Contributor
Contributor

Did you ever find a resolution to this?  We are experiencing the exact same issue between our UCS and VNX 5400 in both of our data centers.  We are on ESXi 5.5 U2.  We have the issue when using a VM Fex VDS, a VMware VDS, or a VMware vSwitch.  We have a mixture of issues as not every host has the exact same issue.  Most of the hosts do not show all of the Port Groups being used, but some show all of them used, but the number of Connected Targets do not correlate that.

David Prows Fort Wayne VMUG Leader
Reply
0 Kudos
lodoss900
Enthusiast
Enthusiast

I had this and it ended up being a port on the switch.  Change the port on the switch and Active/Active

Reply
0 Kudos