General KVM/Proxmox question: ping response for offline (qemu) virtual machines

Hi all,

not a specific OpenSuse question, but as all my systems are OpenSuse, I beg your pardon for starting to ask here.

I’ve a virtual server (VPS) rent from a German hosting company. OS on VPS is OpenSuse Leap 15, but issue probably independent.
As far as I can see when “looking” at the boot process via VNC remotely, Hosting company uses Proxmox to start my VPS. No info about versions and further infrastructure.

When my VPS is offline, e.g. correctly shutdown-ed, a “ping <IPv4>” to this VPS hangs without any response, for hours. Earliest after 6 to 8 hours or so I’m getting the expected ICMP “destination host unreachable” messages, then at once, after very short time.

My questions are:
Is this “hanging” a typical behaviour for proxmox/kvm based virtualized infrastructures, or can this be changed, somehow?
For my understanding, in case of offline, the last router must trigger the “destination host unreachable” message. So in case of virtual machines probably the host system?

The (very nice and helpful) tech support guys from the hosting company have told me, that they don’t get these “hangings”, instead they are getting “network timeout” msgs for their pings after a couple of minutes of offline time. No idea why, as my pings are always hanging, independent of starting point. Tried from my OpenSuse box at home, tried via my android smartphone (via connectBot, shell session for “localhost”), when connected via WLAN to my wired internet connection, and also when connected via mobile data to a different telco provider: Pings are hanging always :frowning:

Any hints would be great!

Thanks in advance,

The questions you pose have mostly to do with your VPS virtual infrastructure and little else, so there’s probably little I can provide you with any certainty.

Regarding your ping response…
When any kind of request “hangs,”
That’s always an indication that the request** was successfully received by something**, and so now your machine is waiting for a response… waiting… waiting…
Until your machine gets tired of waiting and times out.
Note that this is different than if your request was either actively rejected (would produce an immediate response) or if your request was not accepted by anything for instance by mis-routing in which case depending on whether your connection is IPv4 or IPv6, your machine would time out more quickly (IIRC IPv4 typically times out in about a minute or two at most, IPv6 might take much, much longer but will timeout).

If I were to guess,
Your VPS Provider likely has proxy edge devices that are causing your problem.
Note also that if the problem is caused by VPS edge devices, what you see would be different than what your VPS Provider Support might see if they are already behind those edge devices.

The best you can probably do is contact a higher level of your VPS Provider support or forward this post that speculates on the cause of the problem you describe(For instance even low level Tech Support can verify your issue by pinging from outside and not from their own machines)… Of course, once any problem including this is properly understood it’s likely IT can address the issue.