what's the status of your physical switch port where your mgmt vswitch uplink is connected. is it configured with appropriate VLAN trunking, and your workstation from where you are trying to reach to esxi host is also in same network as VLAN 5. Its basically physical to virtual network communication seems to be affected here.
are you vMotion and Mgmt uplinks on same physical switch or different?
You set the new VLAN ID in the Management port group properties?
Did you create the same vLAN in physical switch and made trunk to that VLAN in which you want the communication.
Whenever you separate the traffic on the virtual switch(VLAN) then please do make sure to enable it in physical network.
Thanks for your reply.
There are 3 vSwitch set up for an ESXi Host - 1 x MGMT / 1 x vMotion / 1 x VM. Each vSwitch has 2 NICs connected to different switch for redundancy.
Network Administrator sets up VLAN this morning and writes "I teamed all 6 NICs together, and trunked them on the switch side" and gave me the VLAN ID to be used. He assigned different VLAN to different vSwitch - 5 / 15 and 25.
Since then, I am not able to ping the ESXi Host nor connect it in vCenter.
Both my workstation and vCenter are not connected to VLAN 5 and it is the cause. I would like to know what I should ask for the Network Administrator to make sure that I can manage the ESXi host. Needless to say, I haven't changed anything for the ESXi host in vCenter as I am not able to manage it.
That doesn't sound right. You have 3 different vSwitches with 2 ports on each so you can't team them together at the switch side.
This would have been right if you had a vDS with 6 uplinks and different portgroups per function which you don't.
What you need from network is to configure 3 different teams one per vSwitch and to start with the one that you have the management need to be in access mode so you can get back the connectivity with your ESX box.
Also the vMotion doesn't need to be in trunk mode as you will have only vMotion. The Data (VM) need to be trunk.
We are only using Enterprise Edition and thus we don't get Distributed vSwitch. We are only using standard vSwitch.
Network Administrator says that his set up will provide high speed (Link Aggregation) and Trunking.
Not sure if this is related to your issue but we had an issue a few months ago when we upgraded our ESXi 5.1 hosts to 10Gb connecting to Cisco 4500s.
Previously, we had specified the VLAN ID that the management network connected to under the vSwitch config but when we moved to 10Gb, we could only get it working by omitting the management network VLAN ID from the vSwitch config. It was related to the "native VLAN" on the switch. The switch port set-up on the 4500 switch is:-
switchport trunk native vlan 111
switchport trunk allowed vlan 555-999
switchport mode trunk
storm-control broadcast level 50.00
spanning-tree portfast trunk
This could, however, be attributed to the fact we now have all traffic going over the 10Gb connection (management, storage, vMotion, VM traffic) hence why it is set-up as a trunk port. Obviously not recommended but this was enforced upon us.
Try leaving the VLAN ID off for your management network as attached. May be worth mentioning to your network team to see if this is related to your issue.
Apologies if this has nothing to do with your set-up.
VLANID.JPG 11.3 K
Don't listen to network administrator.
If you have one team with 6 ports at the network device and 3 vswitches with 2 ports in each and different config at the server side how this will work?
Probably your network administrator doesn't know what exactly you have configured at the server or if you told him and he insist then good luck with your network :-)
always make sure you have control over the network too. i think vmware engineers should be network/storage/security administrators too
but sounds like misconfigured VLANS
don't tie the ports together because you said every nic is on a different switch for redundancy so assign the correct VLANs to each port
make sure your vcenter server has access to management vlan and also your workstation of course
i would recommend to just make a sketch of your configuration and sit with your network admin to make the correct config
Unfortunately, we don't get any access to CISCO routers and no documentation for settings.
I have explained to him that there are 3 vSwitch being set up before he set up the switch settings.
He says that the vCenter Server should be able to access the ESXi Host and he has already done his part. It is my responsibility for going to Console, NIC Teaming and makes it work.
In that case although is not recommended or best practice leave only one vSwitch, add all your cards in the team as active, configure all your networks with VLAN ID.
But before you manage to do that and to get access to your server try to login to console and specify the VLAN ID on the management network as per the picture below:
Let me know if it will work.
You are right.
He doesn't listen but finally changes his configuration as follow
Nic0 and Nic5 have been teamed together for Management
Nic1 and Nic6 have been teamed together for vMotion
Nic2 and Nic7 have been teamed together for the VM
May I ask for curiosity
1) NIC Teaming for Physical Switch is for Link Aggregation and Load Balancing ?
2) Does the NIC2 and NIC7 need to be trunked at Physical Switch
BTW, what is the meaning of Trunking ? For what purpose ?