Cloudinit is loading many things, so I wouldn't be surprised it makes the boot longer. Can we check with a default/basic Ubuntu server install and give the timing?

@cyford said in after downgrading xcp host from trail 90% of my vms are not booting..:

@Danp

just want to say after reinstalling all and recreating the all repositories i do have any disk errors or IO issues

I assume that you meant to Include a not in there. 😁

@alpenliebeanj

I know you resolved this a different way but what I did when migrating P2V from a bare metal debian system to xcpng is using clonezilla I transferred the image via the network to the VM which was booted to clonezilla as well. Then I changed the boot option to UEFI in advanced, pressed escape on startup and then proceeded to add a boot menu entry for grub of the Debian system. I would assume the process would have been similar in your case as well to resolve.

Thanks! We didn't test on threadrippers yet, so it's useful info 🙂

@olivierlambert said in Xcp-ng 8.2 no more recognized with Cloudstack since last update:

ill review the new pricing, but xcp-ng / cloud stack documentation is quite lacking. it used to be pretty much plug and play back in the day

It should be plug and play, but as I said, we always been very friendly with CS devs and we even participated to a CS event in last November. If the integration is less good nowadays, you should tell CS people, because our API is stable since it exists. We can't be blamed if it works less well (if it's even confirmed) because that's not on us 🙂

for the lazy 🙂 XOA dfrom sources doesnt have alot of the features needed, like deploying K8s clusters, or templates and thats just an example

So you want turnkey and cheap/free 🤔

no, i want working and affordable, just checked the new pricing model, yes, now it appears affordable. bonus points.

i installed a fresh 8.3 the other day and got XOA on it, tried to deploy a k8s cluster, it failed. only deployed a master did nothing else, second to this i added a ssh-key to debian and installed from a template, could not login to debian, it did seem to work on almaLinux just fine

If it fails and you don't report it (or nobody does), I don't see how we could help 🤔

i have another deployment at home, ill test the k8s again in the lab, as for the ssh keys, i did mentioned it on the channel, just in case its not a bug but i did something wrong.

Yupp, thats me being frustrated.. trying to get something to work so i can move forward.

Please try to be nice, because even if you are frustrated, being not nice won't help to get assistance for free on a community forum 🙂

Okay then try CPU cap at 400 🙂

@Danp
I do not have a power outage or did you recently install XCP updates.
I shutdown the host at the console.
Next day,I power on the host,and that was it.

@kait said in how to disable vm console or make it secondary display device:

@DustinB im using passthrough not vgpu but i could try that anyways

I assumed you were attempting to use passthrough.

@alexredston Hey, sorry I'm a little bit late here. So, with regard of VCPUs - there's a hardcoded limit of 128 actually in XEN hypervisor. Moreover XEN toolstack (when creating a guest) will check that guest VCPU limit is below physical CPUs available on the platform.
Bypassing the VCPU 128 limit will require some rather important adjustements in XEN Hypervisor (basically the restrictions go with IOREQ server from QEMU and how LAPIC id are affected in XEN). So with the next XEN version this limit could potentially be increased (there's an ongoing work on this).

The things you also probably would like to know about this VCPUs limit

not all of the guests can handle this VCPUs number (e.g. Windows will certainely crash) when you gives a VM such a big VCPU number (basically more than 32) the VM can potentially provoke the DoS on the whole platform (this is related how some routines are "serialized" in XEN Hypervisor). So, when you do this - be aware that if your guest is broken, pawned, whatever... your whole platform can potentially become unresponsive.

Finally, I follow an existing topic and configure nut-server on the hypervisor to access the ups via usb.
I know that Olivier is not fully aligned with the fact that the host is modified but IMO it is an acceptable change on my XCP-NG host.

From my perspective, there's literally money on the ground for any virtualization platform to pick up VDI with Intel. The GPU's are affordable and performant for VDI work. They currently work with Openshift and Proxmox is at work on it.

@olivierlambert Do you think it's possible for that support to be added, or are there proprietary code limitations that prevent it?
What would it take to solve that limitation?

We have currently delivered a project which which uses vmware, and the current temporary licenses end in September (could probably be extended a little).
In our cluster we have a total of 10x Nvidia A2, and 3x Nvidia L40 cards.
I would love to move to xcp-ng, as it really seems perfect for us, but the lack of Nvidia vgpu support stops us.

@enes-selcuk The error is telling you that you're host doesn't have enough Memory to actually start the VM.

Is this VM from an OVA and being imported (or has been imported?)

If you have XO or XOLite installed, you can try to reduce the amount of Memory assigned to the VM and start it then.

In case someone ends up here with a similar issue, you also end up at the UEFI shell when UEFI is selected for a BIOS-based image.

Try with Advanced -> Boot firmware changed to bios.

As long as you configure things right like @DustinB said you should be good to go, I don't believe anything will override that and sync to the host.

@ambad4u For sure! Yeah, 1.29 does not work for me either. Going to keep an eye out for any new versions and report back.

For the record, I only have this one unit to test; so there's always the chance mine could be bunk. Luckily these units make it easy to change out the firmware in either direction, so anyone willing to (carefully) test and confirm, would be appreciated.

Been running some benches and it's a fine little machine otherwise.

@high-voltages you also can try xenpm start 10 to gather info for 10s and it will print a sumarry at the end, in my case it had each core average frequency, not sure you'll get it too, but you can give it a shot.

@XCP-ng-JustGreat
We ended up doing the Registry bypass. After that was done it worked great!

We also tried things on our ESXi cluster, but due to licensing, couldn't do the vTPM. Also made sure to mention to my IT Manager about Olivier's 20-minute response to my forum post. 🙂 Very much appreciated - hopefully I can get us to switch over.

@enes-selcuk odd...

Sorry I can't be of more help with this than... when you're VM is starting I think the boot key is the Delete key, maybe there is a keyboard difference?