I may have spoken too soon. The tools were installed (apt install xe-guest-utilities). But both XCP-center and XO from source tell me that the tools are not detected.
Do not install the guest tools from the repositories, it's completely broken, and if you do, then removing it is very difficult. I had to wipe everything and start from zero to get this right. But I can confirm that on a completely vanilla Ubuntu 22.04LT install the guest tools on the ISO that comes with 8.2 work just fine.
As near as I can tell the apt version does not install the xe-linux-distribution.service and run it, and therefore the various daemons won't load.
You can read more about my plight here: https://xcp-ng.org/forum/topic/7494/ubuntu-22-04lts-guest-utiilities?_=1688263726663
I may have spoken too soon. The tools were installed (apt install xe-guest-utilities). But both XCP-center and XO from source tell me that the tools are not detected.
Do not install the guest tools from the repositories, it's completely broken, and if you do, then removing it is very difficult. I had to wipe everything and start from zero to get this right. But I can confirm that on a completely vanilla Ubuntu 22.04LT install the guest tools on the ISO that comes with 8.2 work just fine.
For anyone else finding this topic when researching guest tools for Ubuntu 22.04LTS it would appear that they have added them back into the official repositories, and as of the date on this post they are version 7.20.2
As near as I can tell it works as advertised.
Hi everyone,
I am having an issue with some VMs that run java based applications. Basically whenever Java kick into high gear, the entire VM locks up and has to be forced to restart.
The two applications in questions are Papercut and Minecraft (running in a docker container).
Anyone else seen this or something like it?
EDIT: I should mention that this does not happen when the VM is Ubuntu 18.04, but I am not sure about 20.04 (which I am working on testing now).
EDIT 2: 20.04 LTS does not crash with the Java app installed.
@olivierlambert said in Accidentally set domain memory to 30Gb, system only has 32Gb of RAM:
Edit grub before it boots and reduce the number
Then it will boot, and re-enter the correct amount
Perfect! Thanks a ton, that did the trick.
Quick follow up questions, if I change it in Grub, is that change permanent, or do I still need to change it via the CLI or in Xencenter?
And I believe this is causing the system to not boot. I get a kernel panic with an error about not being able to enable IOMMU.
What's the easiest way to fix this? Or am I looking at a complete re-install/upgrade type fix?
That makes sense. So the way to do it would be to install all the updates, then restart the Master, followed by the rest of the hosts?
If, it might be good if the tooltip for the exclamation triangle was a bit different on the non-master hosts, could say "Reboot the master, then this host to apply updates" vs "Reboot this host to apply updates"
Is there an official way to make suggestions like this? Not necessarily a feature request, more a UI improvement suggestion.
Thanks, I actually tried all of that, it did not help. I then rebooted the pool master (which also had updates pending) and once it came back up that alert from the host in question went away. I guess "reboot the host to apply updates" really means restart all the hosts in the pool to apply updates.
Maybe I could suggest a bit clearer description of that flag?
Also, there is a place in XO that I can go to see exactly what patches are installed and applied to any given host and/or the pool?
So last night I compiles XO from source as I plan to use it at home for my 2 XCP-ng 8.2 servers, but also as an eval for work, since I am thinking about moving to it there from XCP-center.
Everything went well, got it up and running and playing with it, but I have run into an issue that, in my opinion, make XO somewhat suspect as a management platform for XCP.
I installed the patches is suggested, and then a triangle appeared in the hosts list saying I need to restart in order for the patches to be applied, I did this three times, but the alert has not gone away.
Related to this, one host also has a red triangle saying that the time on the host does not match that of the XO system. Which was true, the RTC was set to local time, but XCP thought that was UTC, so it adjusted the system time. I fixed that, now the time is correct. Yet the triangle has not gone away, it's been over twelve hours and a reboot of that host, the red triangle persists.
Did I miss something here? Is there a way to force XO to check again and update? How do I know if the patches have been applied (XCP-center used to be nice and easy for this)?
I also note that none of these warnings/errors are listed under the Healthy dashboard in the alerts section, or in the Overview dashboard.