hi,
you should be able to force stop the VMs by hitting "Stop" on the power menu.
if that is not working, then it might be because you have another "Shutdown" or "Reboot" pending... you can cancel that in the task log (at the bottom).
if none of that works for some reason, the most forceful way of killing the VM without rebooting the host would be to just kill the
you should be able to force stop the VMs by hitting "Stop" on the power menu.
if that is not working, then it might be because you have another "Shutdown" or "Reboot" pending... you can cancel that in the task log (at the bottom).
if none of that works for some reason, the most forceful way of killing the VM without rebooting the host would be to just kill the
kvm process for that VMAttachments
Hi @oguz
Thanks for reply.
its about several years i use now proxmox (as well Esxi : use this one even longer...). I of course try a lot of possbilities each time Including commands etc...
To be admin and cannot even force to stop a VM (no matter what..) it is sooooo .. how to say.... (GRRRRRR.......
!)
I have to reboot so many time that now PVE server do not boot anymore. I ll probably have to check it tmr morning (at office) whats going on.... Its really really annoying.... Never experienced it for the other brand....
Thanks for reply.
its about several years i use now proxmox (as well Esxi : use this one even longer...). I of course try a lot of possbilities each time Including commands etc...
To be admin and cannot even force to stop a VM (no matter what..) it is sooooo .. how to say.... (GRRRRRR.......
!)I have to reboot so many time that now PVE server do not boot anymore. I ll probably have to check it tmr morning (at office) whats going on.... Its really really annoying.... Never experienced it for the other brand....
Last edited:
when the "Stop" job doesn't kill the VM it's probably because there's another pending "Reboot" or "Shutdown" job going on, which you can cancel in the task log at the bottom if you double click it and select "Stop" inside the task (i guess that's a bit confusing and should be rather called "Cancel" or so though...)
as mentioned, you can kill the kvm process if need be. just try:To be admin and cannot even force to stop a VM (no matter what..) it is sooooo .. how to say.... (GRRRRRR.......
Code:
ps aux | grep <VMID> # will show you the kvm process ID running your VMID
kill -9 <PID> # you can now kill it with the process IDbeware that this is equivalent to pulling the plug on a physical machine though
Probably the most annoying thing about proxmox that I have ever seen: shutdown, stop, reset immediately, ALL OF THEM FORCE OPTIONS ARE USELESS, don't work, hang etc...
In my case the dumb thing is stuck with "unable to unmount CDROM", which I have forcefully removed after this, nothing works.
Now dear proxmox people: This is your problem, your bug, if you use virtualmachine-manager on Linux which runs KVM / Qemu VM, ALL of the options work IMMEDIATELY, i.e. right now, force shutdown and force restart are an necessity for all VM hosts: we all know that machines tend to get stuck at the end and not reboot or whatever and you need to push and hold the button. I don't know why this is broken in proxmox.
Otherwise, I'm a big fan of your software, it is really awesome, but being involved in QA, I really don't know why you don't fix obvious bugs like this.
I will now have to restart the whole VM Host, which is absolutely no feasible solution.
In my case the dumb thing is stuck with "unable to unmount CDROM", which I have forcefully removed after this, nothing works.
Now dear proxmox people: This is your problem, your bug, if you use virtualmachine-manager on Linux which runs KVM / Qemu VM, ALL of the options work IMMEDIATELY, i.e. right now, force shutdown and force restart are an necessity for all VM hosts: we all know that machines tend to get stuck at the end and not reboot or whatever and you need to push and hold the button. I don't know why this is broken in proxmox.
Otherwise, I'm a big fan of your software, it is really awesome, but being involved in QA, I really don't know why you don't fix obvious bugs like this.
I will now have to restart the whole VM Host, which is absolutely no feasible solution.
Shutdown takes a long time to time out if the VM does not respond (because QEMU Guest Agent has already shut down or ACPI is ignored) and Stop won't work because Shutdown is still running (and locking the VM). Quite annoying but during Proxmox host shutdown all VMs that do not respond to Shutdown are Stopped eventually. Would be nice if such a command would not wait for the previous ones but interrupt them instead.
Jup, in general a shutdown task should forcefully stop the VM after a timeout triggers (I think its 3 or 5 minutes). But sometime this still won't work. I'm seeing this when shutting down a Win10 VM that want's to install updates while shutting down. But in such a case it still works to stop the shutdown task (double click on the stuck shutdown task at the bottom and click the stop button). This will release the lock and makes it possible to start a new task like a stop or reset. These will then kill the VM.
We need FORCE options that work, always. You as the admin know when you can kill and forcefully remove a VM, no matter what, no timeout, no waiting, no asking stupid questions, just do it and force it.
Thats already the case, you just have to stop the shutdown task before starting a stop/reset task.
That's a bug then. You want the force task to go through, it is normal and expected to first try nicely, then force it. The force task must override the old task, what good is it otherwise?
It is kind of switching on the fire suppression system in a fire, then realize it is so dire that you need to call all the fire department urgently. Now with the logic it is now the fire department will never show up because you first have to cancel the fire suppression system, it will first wait for that to fix the problem. I would think this stop button needs to cancel any shutdown that is already running, this is not the only post, there are so many and most people expect that there is a way to switch off the VM without waiting and without it being able to get stuck. the best way would be to run them parallel, so the first shutdown task will just fail once the VM is off, and in fact it really failed because it did not manage to terminate the VM.
Last edited:
I agree with you. It's not that obvious how to forcefully stop VM. Didn't got that myself at first. Just wanted to make clear that this is still working using the webUI when you know how to do it with just a second more of work.
It is also not clear what the difference between stop and shutdown is. Force shutdown is a better term.
"Force shutdown" would imply that it is doing a shutdown. But "Stop" isn't doing any shutdown, it will just kill the VM like pulling the power cable out of a running physical host.
"Shutdown" will do a shutdown (first gracefully and if that won't work it should stop it with force after a timeout which it sadly not always does).
"Reboot" will do a "Shutdown" and a "Start" after that.
"Reset" will do a "Stop" and then a "Start".
"Shutdown" will do a shutdown (first gracefully and if that won't work it should stop it with force after a timeout which it sadly not always does).
"Reboot" will do a "Shutdown" and a "Start" after that.
"Reset" will do a "Stop" and then a "Start".
Last edited:
No, stop is a term that can have many many meanings, it can also mean suspend. Only the term force says it all, you force things when they otherwise don't work and/or are broken.
Stop is actually dangerous, because it does not clearly say what it does and the word is defined differently on different systems. But I don't know why we are having this discussion, this is clear: you either use clear terms like force that are unambiguous or you use ambiguous ones and expect admins to RTFM, and such a simple concept does not need RTFM, right now it is no intuitive GUI or you risk them making mistakes by force killing a VM while they just wanted to stop the VM (which is just another words for shutdown or poweroff) because they understood their meaning of the word stop. I know now how to get around it, but this posts will be asked over and over and over and you will waste your users time by not explaining it correctly. Explaining it in the GUI that is.
Shutdown is a widely used term for, well graceful shutdown, of an OS showed as such in *all* major OS. "force shutdown" would be much more confusing for people as while it implies that the shutdown is forced but not when that's the case, so it really isn't conveying the (hard) stop semantic.
Stop is quite clear, especially in the context of PVE, where it's a sub menu besides the commonly used (graceful) shutdown button, where the latter is even the default one, which we always try for the safer and more commonly wanted one.
Yes sure, UX can always be improved, but nobody thinks 1:1 the same, so it will never be perfectly safe for everybody, so rather one should just get some basic training, especially for standard things like working with the VM power state.
The thing to do for improving UX here is not to rename the buttons, that will just confuse some other users with perception, but rather add a respective "this will immediately hard stop the VM, danger for data loss, yadda yadda" hint to the prompt.
W.r.t. stop not stopping: this stems mostly from the Linux Kernel's uninterruptible sleep state (the "D" state) that processes may be in, e.g. if NFS connection is down and VM is waiting on IO, those processes just cannot be stopped at all while in that state, only thing is (with luck) a reboot or maybe even cutting power, so this is independent of PVE and some devs of upstream kernel are pondering better semantics, or at least reduction of cases where the "D" state can happen in the first place.
What we'd like to change in the shutdown/stop procedure though is that a later stop trumps an earlier, still running, shutdown, basically overruling it. That makes sense to have for convenience.
Stop is quite clear, especially in the context of PVE, where it's a sub menu besides the commonly used (graceful) shutdown button, where the latter is even the default one, which we always try for the safer and more commonly wanted one.
I'm not sure why some admins/it workers think they have the right to just blindly work with important systems without actually reading manuals, testing the new system/software/ui that they aren't used to yet and getting some basic knowledge of that. In every other workplace/domain this would be a no-go, think of a heavy-duty machine operator just starting a new machine without any experience nor training with it, mashing buttons, destroying things and blaming the maker for a (for them) unclear button label.
Yes sure, UX can always be improved, but nobody thinks 1:1 the same, so it will never be perfectly safe for everybody, so rather one should just get some basic training, especially for standard things like working with the VM power state.
The thing to do for improving UX here is not to rename the buttons, that will just confuse some other users with perception, but rather add a respective "this will immediately hard stop the VM, danger for data loss, yadda yadda" hint to the prompt.
W.r.t. stop not stopping: this stems mostly from the Linux Kernel's uninterruptible sleep state (the "D" state) that processes may be in, e.g. if NFS connection is down and VM is waiting on IO, those processes just cannot be stopped at all while in that state, only thing is (with luck) a reboot or maybe even cutting power, so this is independent of PVE and some devs of upstream kernel are pondering better semantics, or at least reduction of cases where the "D" state can happen in the first place.
What we'd like to change in the shutdown/stop procedure though is that a later stop trumps an earlier, still running, shutdown, basically overruling it. That makes sense to have for convenience.
Last edited:
Where we are on the subject:
Why is the "Bulk Stop" named so, while it actually does (fortunately) not stop the VMs/LXCs, but rather shuts them down?
"Bulk Shutdown" with the appropriate icon would be correct.
Currently the naming is inconsistent and basically wrong or do I miss something?
Why is the "Bulk Stop" named so, while it actually does (fortunately) not stop the VMs/LXCs, but rather shuts them down?
"Bulk Shutdown" with the appropriate icon would be correct.
Currently the naming is inconsistent and basically wrong or do I miss something?
Well, bulk stop IS a stop though, it's just a combined one with timeout and force stop happening after the time limit got exceeded.
https://git.proxmox.com/?p=pve-mana...93a40f02fa9c4fd3e3f5e3c5cab5fb6;hb=HEAD#l1885
Ah, okay. Thank you for the explanation.
Maybe I am the only one who thinks so, but I had to google first what it actually does, because I was scared that it does directly (force) stop the VMs/LXCs, like the per machine "Stop"-button does.
