VMware Cloud Community
HDSNDL
Contributor
Contributor
Jump to solution

Cannot add physical NIC's to vSwitch

I built this ESXi 6.5 lab machine (6.5.0 (Build 4887370) with 2x physical NIC's and it's been fine. I wired up and verified the second NIC pair recently. They won't add to the machine.

vmnic0 and vmnic1 were the original 2. I added vmnic2 and vmnic3 today.

3.png

6.png

I tried the CLI method:

[root@vmware65hv1:~] esxcli network vswitch standard uplink add --uplink-name=vmnic2 --vswitch-name=vSwitch0

Unable to Set: Sysinfo error: BusySee VMkernel log for details.

[root@vmware65hv1:~] esxcli network vswitch standard uplink add --uplink-name=vmnic3 --vswitch-name=vSwitch0

Unable to Set: Sysinfo error: BusySee VMkernel log for details.

[root@vmware65hv1:~] tail /var/log/vmkernel.log

2019-05-23T15:55:57.514Z cpu16:116571)NetSched: 628: vmnic2-0-tx: worldID = 116571 exits

2019-05-23T15:55:57.514Z cpu1:65629)Uplink: 9893: enabled port 0x4000002 with mac d0:67:e5:e7:5c:8f

<6>bnx2 0000:02:00.1: vmnic3: NIC Copper Link is Up, 1000 Mbps full duplex, receive & transmit flow control ON

2019-05-23T15:56:02.513Z cpu8:65629)NetPort: 1879: disabled port 0x3000002

2019-05-23T15:56:02.514Z cpu5:116572)NetSched: 628: vmnic3-0-tx: worldID = 116572 exits

2019-05-23T15:56:02.514Z cpu8:65629)Uplink: 9893: enabled port 0x3000002 with mac d0:67:e5:e7:5c:91

2019-05-23T15:56:14.128Z cpu18:66989 opID=e56c7b71)World: 12230: VC opID esxcli-d4-7a1e maps to vmkernel opID e56c7b71

2019-05-23T15:56:14.128Z cpu18:66989 opID=e56c7b71)NetTeam: 715: Failed to register uplink vmnic2: Busy

2019-05-23T15:56:19.487Z cpu5:67726 opID=9c0d5f)World: 12230: VC opID esxcli-3b-7a29 maps to vmkernel opID 9c0d5f

2019-05-23T15:56:19.487Z cpu5:67726 opID=9c0d5f)NetTeam: 715: Failed to register uplink vmnic3: Busy

7.png

Any idea what may be the cause of this?

Thank you in advance.

0 Kudos
1 Solution

Accepted Solutions
daphnissov
Immortal
Immortal
Jump to solution

And you're certain you have none of these other NICs in a team on some vSwitch out there?

View solution in original post

12 Replies
daphnissov
Immortal
Immortal
Jump to solution

Before going further, I'd just say that you appear to be using 6.5 GA which is horribly outdated as far as patches are concerned–one of which might address the issue. So bring it up-to-date first of all and see what that gets.

0 Kudos
HDSNDL
Contributor
Contributor
Jump to solution

Thank you, I appreciate that heads-up. I just finished upgrading to 6.7.0 Update 2 (Build 13644319), and am experiencing the same issue. Any ideas?

0 Kudos
daphnissov
Immortal
Immortal
Jump to solution

I didn't necessarily mean upgrade; a simple update would have been fine, but anyhow. Need to see more info and detail on what you're experiencing.

0 Kudos
HDSNDL
Contributor
Contributor
Jump to solution

What more info would you like to see, other than what was posted originally?

Basically, it has 4x 1GbE onboard connections. I originally installed it with 2 connected. Earlier today, I wired up and verified I have link on the new 2. ESXi sees them and they are enabled. They are connected to a Juniper EX4200. I do have link on all 4 ports, and they are active

0 Kudos
daphnissov
Immortal
Immortal
Jump to solution

And you're certain you have none of these other NICs in a team on some vSwitch out there?

HDSNDL
Contributor
Contributor
Jump to solution

100% sure. They've never been used up until today.

0 Kudos
daphnissov
Immortal
Immortal
Jump to solution

So here's what I'd try:

  1. Disconnect those unused NICs.
  2. Shutdown ESXi host.
  3. Pull that 2-port card completely.
  4. Power up ESXi host. Ensure those PCI devices are no longer showing up.
  5. Shutdown ESXi host again.
  6. Re-install 2-port card.
  7. Power up ESXi host last time.
  8. Ensure they're showing again.
  9. Cable NICs so they're showing connected.
  10. Try to add to new or existing vSwitch.
0 Kudos
HDSNDL
Contributor
Contributor
Jump to solution

I wish I could, but those 4 NICs are onboard, integrated NICs. Server hardware is a Dell PE R710.  I'll try updating BIOS and a few other things and will report back when I do. Thank you so very much for the help.

0 Kudos
daphnissov
Immortal
Immortal
Jump to solution

Another thing to try is disable those in the BIOS and reboot (substitute from above).

0 Kudos
HDSNDL
Contributor
Contributor
Jump to solution

Thank you daphnissov! You were EXACTLY right. I double-checked a couple other vSwitches that are used for backend cluster communication amongst some vm nodes, and found those two interfaces were bound to them, where they didn't need to be, since it's a non-routable network for private cluster multicast traffic. Thank you!!!!! 

0 Kudos
daphnissov
Immortal
Immortal
Jump to solution

Good, I was almost positive that was the situation. Glad you found them.

0 Kudos
UCGuy1
Contributor
Contributor
Jump to solution

Just as an update to this thread, depending upon if you've upgraded to ESXi 7.x, you can go "directly" to the ESXi server, under the menu "Network Restore Options," you can create the vswitch, portgroup and uplink in this one spot.  I just happened to try this method as a last-ditch effort and it worked successfully.

0 Kudos