XCP-ng 8.3 betas and RCs feedback 🚀
-
Ah thanks, great feedback
-
@olivierlambert Thank you. Install went smoothly and I'm in the process of transferring all of the VMs from the backup server back onto my main workhorse.
I just wish I remembered an old troubleshooting tip:
When something odd is going on, reset the bios to defaults (or something along those lines).I did figure out a way to run the installer from efi shell; never done that before. Wish I thought of it sooner as well (no telling how many times I've rebooted this system trying to troubleshoot things.
-
@eb-xcp said in XCP-ng 8.3 betas and RCs feedback
:Edit: Confirmed; after enabling execution disable option within bios, installer booted without issues and the install is currently ongoing.
That is a bug. Xen is supposed to be able to detect this case and re-activate NX on it's own.
For the EFI path in your screenshot, that one doesn't have logic to re-activate. IIRC, we weren't sure whether it was needed, because surely an EFI system wasn't still using Pentium4 compatibility. Clearly some wrong reasoning, and it's fairly easy to adjust.
However, fixing that path wont fix the normal MB2 path, which does have logic to reactivate and should have been able to cope fine.
What system do you have?
-
@andyhhp thanks, seems issue was most visible within RC1 install. 8.2 worked well with NX disabled (I guess by activating it on it's own), beta 2 would install, but not boot (have not tried beta 1), and RC1 would fail seemingly silently.
The efi path is not to activate the NX setting, it was just me getting to the cd directory with the bootx64.efi launcher for the installer. Running it this way, rather than allowing BIOS to boot into CD directly is what allowed me to see the output of the installer prior to exit: "This build of Xen requires NX support". Normally the next boot option would take over and this text would not be seen.
My setup is Supermicro X10SRM-TF motherboard with E5-2699 V4 processor, latest bios, UEFI boot.
I don't even remember when I disabled NX support, perhaps couple of years ago when playing around with increasing single core performance. -
@eb-xcp said in XCP-ng 8.3 betas and RCs feedback
:beta 2 would install, but not boot
Did you use the network to install the packages? I would expect that when getting them from the ISO you should have the same behavior on first boot than when booting the ISO (but obviously, once the packages get upgrade later, that would bring the problem too)
-
@yann Never got that far, 8.3 beta 2 would not boot after the install.
Difference with RC1 is that the beta 2 installer itself did not check, or could re-enable NX support, but the actual boot of beta 2 would fail similar to RC1 installer (likely due to NX support being disabled).
So I could install, but not boot, the beta 2 onto the system. My previous posts don't provide much detail, but here is how it went: https://xcp-ng.org/forum/post/80208 -
Forgot to mention that I did try installing RC1 via legacy bios onto the system (installer would work in that case), but I could not get the system to recognize boot disk (NVME) after the install (installer itself saw and could use said disk).
Ended up discovering issue with EFI installer prior to fixing the legacy path as I've not spent much time on that since legacy bios would not be supported down the road. -
@eb-xcp said in XCP-ng 8.3 betas and RCs feedback
:@yann Never got that far, 8.3 beta 2 would not boot after the install.
Roger that, but when installing beta2, did you select "local" or "http" as package source?
-
@yann Good catch. I tend to use net install when available, as it is faster than emulated iso on supermicro motherboards. When trying beta 2 install, I did use the network source by using xcp-ng-8.3.0-beta2-netinstall.iso.
To test things further, I used my backup server (Supermicro X10DRT-H, with E5-2630 V3) and disabled execute disable bit. Verified that 8.3 RC1 install gives the same error "Xen requires NX support" and ran beta 2 install from local media. This time 8.3 beta 2 was installed without issues and after a reboot I verified that execute was still disabled in the bios.
When using RC1 installer:
Beta 2 installer with local media this time:
-
@stormi said in XCP-ng 8.3 betas and RCs feedback
:@NoHeadroom said in XCP-ng 8.3 betas and RCs feedback
:Anyway, it's better to have an option b) if option a) fails
As you wish, but I won't support option b) in any case, so I prefer to mention it
Yesterday I upgraded my 8.2 server by downloading the 8.3 rc1 iso file onto a USB stick. In general, virtual machines, iscsi storage, etc. everything worked without problems, but there is an interesting situation. Before the upgrade, the server was connecting to the internet without problems. For example, when I ping Google, I get a response, but now when I ping, it finds the IP address from DNS, but there is no response. I checked the gateway and DNS settings many times and did not see any problem. In this case, I could not install XOA either. Has anyone experienced such a problem?
-
@altikardes
I assume there isn't a VLAN configured, isn't it?Anyway, sometimes it's better to do the "Emergengy Network Reset" (under "Network and Management Interface") and start to reconfig your network interfaces from bottom up.
-
Yes, there is no vlan configuration. There was no problem with internet connection before the upgrade. I set the ip configuration as dhcp but the result is the same. I reset the network settings with "Emergency Network Reset" from the xsconsole section and made the necessary settings after the system was restarted but the result still did not change. (I tried both static and dynamic ip.)
-
@altikardes
what's the contens of /etc/resolv.conf ? -
Hello again, the problem was related to a rule change on the firewall. I didn't feel the need to check it because it coincided with the time I upgraded. So everything is working now. Sorry for the false alarm.
-
Good news then
-
I have also successfully run a test install of RC1 on my lab. Everything seems to be working fine. The TLS issue also seems to go away (will also put a separate post on the topic concerned).
Good Job guys !
One thing I noticed :
I seem to get an error message
no crontab for rootwhile running
crontab -ethe first time.Is that normal? Is it ok to run a cron job for root ?
-
@gb-123 said in XCP-ng 8.3 betas and RCs feedback
:I have also successfully run a test install of RC1 on my lab. Everything seems to be working fine. The TLS issue also seems to go away (will also put a separate post on the topic concerned).
Good Job guys !
One thing I noticed :
I seem to get an error message
no crontab for rootwhile running
crontab -ethe first time.Is that normal? Is it ok to run a cron job for root ?
Could be very well happen if you have a fresh install and no cronjobs in the crontab. Most distributions would put corn jobs in /etc/cron.d anyways and so does xcp-ng too. And the users crontab then will be created on the first edit.
-
@gb-123 said in XCP-ng 8.3 betas and RCs feedback
:Is that normal? Is it ok to run a cron job for root ?
I´d advise to rather use system crontabs in
/etc/cron.d/, one dedicated to each topic you have (but that's rather unrelated to XCP-ng itself, just standard Linux sysadmin practice). -
Thanks for the tip !
Anyway to delete the crontab created for root ? Or it is not required (since it is blank anyways) ?(I should have just usedcrontab -ir)I also noticed that I get an option to "Upgrade" to RC1 from beta 2. This should technically not happen since I installed beta 2 in Legacy BIOS mode and am now installing RC1 in UEFI mode since the former would be deprecated.
Is that an error in installer or upgrade is possible ?
-
@gb-123 said in XCP-ng 8.3 betas and RCs feedback
:I also noticed that I get an option to "Upgrade" to RC1 from beta 2. This should technically not happen since I installed beta 2 in Legacy BIOS mode and am now installing RC1 in UEFI mode since the former would be deprecated.
Is that an error in installer or upgrade is possible ?
It's not exactly an error, but it's not very good either. This would fail at some point so don't attempt it. We suggested to add the detection of such situations, upstream, but neither them nor us had the time to implement it: https://github.com/xenserver/host-installer/issues/11
