@stormi Microcode updated on affected Gen11 i7. Running normally.

@gduperrey I jumped in all the way by mistake... I updated a wrong host, so I just did them all. Older AMD, Intel E3/E5, NUC11, etc. So far, so good. Add/migrate/backup/etc VMs are working as usual. Good for guest tools too, but mine are mostly Debian 7-11. Stuff is as usual so far.

I agree and I'll say it again, Thanks! It's not just Linux/Xen stuff. It is XCP-ng and XO that make everything work as a cohesive vertical open-source solution (some nice buzz words). Thanks to the Vates team and community that have built and support it. I look forward to ongoing continuous improvement and innovation!

@bleader Installed and running.

@bleader Updates running on several old and new intel machines (including microcode update). Working fine so far. Rolling Pool Reboot is a helpful feature.

@bleader I installed it on a bunch of busy hosts. All are fine, but none used PCI passthrough. The Rolling Pool Reboot in XO was very helpful.

@julien-f @olivierlambert

Thanks for the ability to delete XO task logs feature! (XO commit f6e6e)

@gduperrey I have several hosts updated and running. I'm happy to see 8.3 updates on parity with 8.2.

@stormi Installed on several test and pre-production machines.

@bleader I installed it on many 8.2 machines. On one I did get a warning:

  Cleanup    : xen-libs-4.13.5-9.40.3.xcpng8.2.x86_64                     14/14
warning: %posttrans(microcode_ctl-2:2.1-26.xs29.5.xcpng8.2.x86_64) scriptlet failed, exit status 1
Non-fatal POSTTRANS scriptlet failure in rpm package 2:microcode_ctl-2.1-26.xs29.5.xcpng8.2.x86_64
  Verifying  : xcp-ng-release-8.2.1-13.x86_64                              1/14

But it did not seem to have any effect. Nothing extra in the yum.log

@LuisRiveraSig Since the device is not need for Xen Console or Dom0 then you can try a PCI pass-thru to the VM or use a USB device and pass it to the VM and just use the needed driver in the VM directly.

@LuisRiveraSig What are you trying to use the serial ports for?

• XCP console?
• Dom0 Serial port (ie. UPS monitor)?
• DomU/VM?

@plaidypus I am not 100% sure what the correct answer is for the default XCP configuration. I think the basic answer is: no. Xen/XCP does not care what cores it uses for your VM. So on average your performance will be a little worse than not crossing NUMA nodes but better than always interleaving NUMA nodes. Some systems will be better/worse than others.

Xen/XCP hypervisor does have a NUMA aware scheduler. There are two basic modes, one is CPU hard pinning where you specify which cores a VM (domain) uses. This would force the VM to use only the cores it is assigned. The other is to let Xen/XCP do its own work where it tries to schedule core use of a VM (domain) on a single CPU pool. The problem with this is the default config is to put all cores (and HT) in a single default pool. There are some options to try and enable best-effort NUMA assignment but I believe it is not set that way by default.

You can configure CPUs of a NUMA node into an individual pool (see below). A VM can be set for an affinity for a single pool (soft CPU pinning). This would keep most of the work on that single node as you want.

The links listed before to have good information about NUMA and CPU pinning. Below are some more:

Here is an older link about Xen on NUMA machines.
Here is a link about Xen CPU pools.
Here is a link about performance improvements on an AMD EPYC CPU (mostly related to AMD cache design).

There are also APIs in the guest tools to allow the VM to request resources based on NUMA nodes.

If you start hard limiting VMs where/how they can run you may break migration and HA for your XCP pool.

@plaidypus Yes, it just works. The guest VM does not see NUMA info and just sees one node even if the config is set for multi socket.

You can't assign more cores than actual threads but you can assign as many as you have to one VM. There will be some performance penalty as the cores in use may use memory from another node.

If you have dual 16 core CPU with HT enabled then you could assign 64 cores to a VM.

@LuisRiveraSig What chipset is used for the serial ports? My guess is that it's a special MOXA serial chipset (non-16550) and requires a special driver that XCP/Xen does not natively support.

@jimfini I'm happy to help if you want to try it. After driver install/reboot, did you refresh your network interfaces on XO? You can look at dmesg on XCP to see the driver loading. Run ip link to see the interfaces.

@gduperrey I have several hosts updated and running. I'm happy to see 8.3 updates on parity with 8.2.

@gduperrey Updates installed and working on busy hosts/pools.

@jimfini I think that USB ethernet is using the RTL8156 chipset.

Please try adding my r8152-module-alt USB driver and see if it works: download page.

USB ethernets are not the best devices for XCP.

Let the community know if this driver works for you as it's new and has not been well tested.

@kagbasi-ngc What OS is on VM that boot to the EFI Shell?