Physical NIC vmnic8 linkstate is down

[RobRunVirt]

Hi,

I have two network connections that aren't used any more. The Event description: Physical NIC vmnic8 linkstate is down was valid, but isn't anymore.
The alert is still comming up in Veeam One monitoring. How do I configure it not to monitor the decommisioned network ports?

Thanks in advance.

Kind regards,
RobRunVirt

[Shestakov]

Hi Roberto and welcome to the community!
You can exclude individual hosts, but not ports.
However, there is a workaround. In event text field you can specify name of networks you want to monitor. All other networks will be filtered out.
Thanks!

[RobRunVirt]

Thanks for your reply. I think that won't be handy in the long run if it is on a higher level object, correct me if I am wrong
There must be a different approach. Maybe removing from a DB and re-discover the ESXi host.

[Shestakov]

Right, if you have several objects where you want to monitor different networks set, you need to create(clone) an alarm for each host.

There must be a different approach. Maybe removing from a DB and re-discover the ESXi host.

Do you mean completely remove information about unused networks? What if it's not used temporary?
Thanks!

[mviel]

We are having the Same Problem ( in our Case we have 2 Ports on a Daughterboard with 4 Ports which we dont use ) , my Current Customer Support Engineer told me to open a Feature Request as the Question isnt solvable through the Monitoring as VeeamOne has now way of filtering those Infos out or handle it differently .

Regards

[Shestakov]

Thanks for the request, Markus.
Have you tried the workarounds described above?

[ChrisGundry]

We have the same issue. When we restart an ESXi host it alerts that one or two NICs are down. These NICs are not in use, but were at some point in the past. All production NICs are online, the ones it flags are NICs that are not plugged in and were not plugged in at any point recently.

We can't exclude hosts or networks from monitoring, those NICs don't have an associated network and we need the hosts monitored for other alerts.

We would also like to request a better way to resolve this, by excluding NICs, or otherwise understand why it alerts on these particular NICs only. Ie our hosts will alert on NIC2 but not NIC0/1/4 which are also not in use/connected...

[Vitaliy S.]

Chris, do you see any events related to these "not used" NICs in the vSphere Client when the ESXi host is restarted?

[ChrisGundry]

Sorry, I am not sure. I will make a note to check next time I get chance to reboot a host.

[Vitaliy S.]

If you see these events (failed ones) then it is clear why Veeam ONE triggers its alarms, but we will think on how to address this situation in our next releases. Thanks!

[wayne7215]

Same here now with the current Veeam ONE build.
Error on a ESXi 6.7 Host with:
Event description: Physical NIC vmnic3 linkstate is down.

We have 6 physical NIC on this host and Nic0 and Nic3 aren't used and have no cable connected. So why is Veeam ONE throwing such an Error for just one of those unused NIC's? On another Host we even have 4 unused NIC's and never have got any similar Error.

[wishr]

Hi Wayne,

We simply catch events that are thrown by vSphere. Could you please make sure they are present there, as Vitaliy described above?

Thanks

[ChrisGundry]

We still get this every time we reboot a host.

We found that we do get ESXi events for the NICs, but we get events for ALL the NICs, but yet VeeamONE is only alerting for one of the NICs.

Example host has vmnic0-11. We only use the 10Gb ones which are vmnic8-11, so vmnic0-7 are NOT in use, never have been.

When we reboot the host, we see these events:
https://imgur.com/a/K232VIl

But yet it only alerted for vmnic7 and you can see on previous occasions it alerts on vmnic2 both times.
https://imgur.com/a/tyAyYuQ

This is the same story each time we reboot a host and the vmnic might vary, but it will always be a single nic from each host that alerts, and one that is not actually in use.

I have two issues:
1. Why is there no easy mechanism to exclude these NICs? Perhaps there is a way to do it, but it would likely involve disabling the default alarm, then setting up new alarms on per-host basis, which is very messy. We would like a tick box type setup on a per-host basis. Think of it like excluding disks on a VM backup.
2. Why is VeeamONE only alerting for a single down vmnic, vmnic0-7 are down, so why is it only alerting on one of them each time?

There is clearly an issue here as several people have reported the same behaviour/issue. I know that Veeam only reports what ESXi sends, but it clearly isn't reporting ALL the NICs as down, so something is being done within VeeamONE to only cause it to alert on a single NIC... Either way, we need better logic and better options from VeeamONE, make it more logical and intuitive for us to use to do what we need it to do.

[wishr]

Hi Chris,

Looks indeed strange. May I ask you to raise a support case so we could collect all the technical information and share your case ID with us?

Thanks

[ChrisGundry]

If I get time I will, but we have around 6 tickets open with Veeam at the moment for various issues so will need to resolve all of them before we can open more!

[billcouper]

Since upgrading to VeeamONE 11 we are also seeing these vmnic* linkstate warnings. In our case, the hosts are not rebooted at all. These warnings just randomly appear. All of our servers have many NICs (vmnic0-7) and we are only using 2 links per host. So hosts have 6 vmnic's that have NEVER been connected, yet we see the alarm for just one vmnic per host. It happens with one host only each time. Not always the same host. Not always the same vmnic. More options to exclude individual vmnic would be great... right-click the alarm, exclude for: this network connection.... please add this feature!

[Vitaliy S.]

Thanks for your FR, Bill! We will discuss it when planning the next releases.

[jeffshead]

Since upgrading to VeeamONE 11 we are also seeing these vmnic* linkstate warnings. In our case, the hosts are not rebooted at all. These warnings just randomly appear. All of our servers have many NICs (vmnic0-7) and we are only using 2 links per host. So hosts have 6 vmnic's that have NEVER been connected, yet we see the alarm for just one vmnic per host. It happens with one host only each time. Not always the same host. Not always the same vmnic. More options to exclude individual vmnic would be great... right-click the alarm, exclude for: this network connection.... please add this feature!

Just want to report that I too have encountered this issue since upgrading to v11.

[wishr]

Hi Jeff,

I've asked our QC team to try to reproduce it again, but may I ask you to open a support case and share its ID so we could collect all the technical details from you in case we need them?

Update: we were able to reproduce the issue and are discussing potential changes that should fix it. I'll share more information here once I have it.

Thanks

[kirov]

I'm just on trial version on VEEAM One because we are looking for a monitoring tool, and yes, I have a lot of linkstate is down alarms of vmnics that are not in use. Glad to see that I was not the only one, I checked the ESXi and did'nt find any alarm/event there, those nics are not in use.

[wishr]

Hi Kirov,

We are planning to allow excluding the nics in the alarm settings in future product versions. Would that work for you?

Thanks

[toma1269]

VeeamOne v12, this is still an issue for us. Is there still no way to un-monitor unused NIC ports (aside from the clunky workarounds mentioned already)?

[jorgedlcruz]

Hello Tom,
Are we talking about the ESXi alarm Host NIC link status? I can see there are exclusions there:


Let me know if that is not the alarm we are talking about.

Thanks!

[Brian_Larsen]

Hi Jorge

Can you provide an example of how you would exclude multiple NICs on multiple Hosts?
So for instance if you have:
Host 1: Nic 1 and Nic 2 not being used, exclude them in Veeam ONE to avoid false positive alarm
Host 2: Nic 3, Nic 4 and Nic 5 not being used, exclude them in Veeam ONE to avoid false positive alarm
Ect...

I have a hard time figuring out the syntax.

Thanks

[jorgedlcruz]

Hello Brian,
If they are different nics what you are after, on different Hosts, you will need to create different alarms, you can clone them, and on the very next step Assignment > Assign that specific alarm with your NICS to the host that is relevant.

Not perfect, but easy to do.

Thanks!

[Brian_Larsen]

Hi Jorge

Ok, thanks.

Can you provide an example of the syntax for the exclude NICs rule? What do you actually type in the white bar?
Do you just type NIC1, NIC2, or?

Thanks

[jorgedlcruz]

Hello Brian,
I will give it a try on the lab and come with the final answer. Thank you

[jorgedlcruz]

Hello, either commas, spaces or ; will work here.

Give it a try as mentioned adjusting the assignments and please let us know

[ItsMeThatsAll]

Hi
Any definitive solution (or workaround)?

[jorgedlcruz]

Hello, we have already shared a workaround.
You create a rule per host, or group of hosts with similar nic distribution, and exclude what is not relevant for you.

Let us know