Alert for network connectivity

[lando_uk]

Hi

I want to know when any of my VM's lose network connectivity for any reason (OS crash, NIC disabled, wrong VLAN, physical switch/gateway failure) we used to use SNMP or basic ping for this, but I'd like to move away from those tools and rely on Veeam One for everything if possible. It would be good if the actual Veeam One server also pinged the VM's that it can see rather than just collecting alerts from the hosts.

I would like a single window to see if everything is green, as right now things may look green on Veeam One, but some of my VM's might not actually be connected and working.

Thanks

[Vitaliy S.]

Hi Mark,

Thank you for your suggestion, but based on my experience most companies have ping prohibited by firewall rules. That said, currently the only way to capture network connectivity issues is either to have an agent inside the guest OS or track VMware or Hyper-V events initiated by a host.

Anyway, we will discuss this FR internally, and will try to find other ways of monitoring VMs network connection state. Thanks!

[lando_uk]

Yes, Server networks would have rules in place, but I'm guessing most server-to-server traffic, particularly ICMP would mostly be allowed or easily added. Pinging hundreds of machines from the Veeam One server could create more traffic, but no different to what a SCOM server or any of those other monitoring tools would do. It's really just the one thing Veeam One is missing meaning we have to still use other tools for this type of basic functionality.

[Vitaliy S.]

Thanks for the feedback, maybe we will be able to add this an option down the road (unless we find a better solution to detect this).

[FranceLink]

Hi,

Is there any news from here ?
Juste installed Update 3, no ping options, neither guest service monitor

[Shestakov]

Hi Gwenaël,
Indeed ping option is not there yet and as was answered in the parallel thread, application-level monitoring is under development. Hopefully it will be released in 2018.
Thanks!

[FranceLink]

Yes I answered it, keep other thread thanks

[dali@iae.nl]

I guess you can use the alarm "VM power status" (with the host icon) which is Disabled by default.
On the Cluster Level right click, Alarms, Define Alarm, click VM power status.
You will get Error alarms for the VM's who are in your off state (probably because you want some VM's normally down). Right click them, Exclude, For object: XXXXX

[RubinCompServ]

The problem with using "VM power status" is when the VM is powered on but has no activity on the NIC. I'm finding issues with one of my servers suddenly losing network access (yellow exclamation point on network icon) with no rhyme or reason. Going into the settings of the VM and disconnecting/reconnecting the virtual NIC resolves the issue. The problem is that I don't find out about it until customers can't access their backups anymore. I would think that, with the VeeamONE Agent installed on this server, VeeamONE should have a way of saying that it lost contact with that machine.

[jorgedlcruz]

Understood the problematic now, can you please check on the affected VM, under VONE monitor, if you click in tasks and events tab, after the performance one, can you see anything? Do you see this warning that vCenter triggers?

If you do, right click, create alarm and follow the usual steps.

Please let me know

[jorgedlcruz]

Hello,
The trick will be if what it causes the VM Network Disconnect, the yellow warning, triggers a vmware event or not. If it does, and we collect it, you can easily create an alarm for that, with your own text, etc.


If it doesn't trigger a VMware event, you will need a third party monitoring script for this purpose.

[RubinCompServ]

@jorgedlcruz

The problem is that it doesn't trigger any type of warning in VMware. As far is VMware is aware, everything is working properly. Considering that there is a VeeamONE agent installed on the machine, we expected that VeeamONE would be aware that it couldn't communicate with the machine (especially based on Vitaliy's comment 7 years ago suggesting that that would be the solution).