Error adding host to a pool

[Usuario Sistemas]

Good morning. ?

When I try to add a host to a pool I get the following error:

You are attempting to add the server 'test' to a pool that is using newer CPUs.

VMs starting on the pool in future will only use the reduced set of CPU features common to all the servers in the pool. VMs already running in the pool will not be able to migrate to the new server until they are restarted.

Do you want to do this?

The problem is that the CPUs used are exactly the same even when checking the microcode matches between the CPUs used and those already present in the pool.

Do you know any solution to this problem?

 

Thanks in advance! ?

[Alan Lantz]

Is the firmware the same as well  ?

 

--Alan--

 

[Usuario Sistemas]

18 hours ago, Alan Lantz said:

Is the firmware the same as well  ?

 

--Alan--

 

 

Good morning Alan. First, thank you for your reply.
Yes, all firmware are up-to-date and up to par with the other hosts in the pool.

 

The information at Dell OMSA level is exactly the same. The CPUs are the following:

 

 

Greetings.

Edited October 15, 2021 by Usuario Sistemas
Added CPU info

[Alan Lantz]

what is the output of xe host-cpu-info of both systems ? It could just be some difference 

enabled/disabled in bios as well.

 

--Alan--

 

[Tobias Kreidl]

There must be a BIOS setting that creates a different CPU mask, as Alan stated above, if the model and firmware are identical.

 

-=Tobias

[Usuario Sistemas]

On 10/15/2021 at 3:39 PM, Alan Lantz said:

what is the output of xe host-cpu-info of both systems ? It could just be some difference 

enabled/disabled in bios as well.

 

--Alan--

 

 

Hello.

 

The pool master's xe host-cpu-info is as follows:

flags: fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush acpi mmx fxsr sse sse2 ht syscall nx lm constant_tsc arch_perfmon rep_good nopl nonstop_tsc pni pclmulqdq monitor est ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm ida arat dtherm
features: 029ee3ff-bfebfbff-00000001-2c100800
features_pv: 17c9cbf5-82b82203-2191cbf5-00000003-00000000-00000000-00000000-00000000-00001000-8c000000
features_hvm: 17cbfbff-82ba2223-2d93fbff-00000003-00000000-00000000-00000000-00000000-00001000-9c000000

 

The one of the host I want to add to the pool has the following:

flags: fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush acpi mmx fxsr sse sse2 ht syscall nx lm constant_tsc arch_perfmon rep_good nopl nonstop_tsc pni pclmulqdq monitor est ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm ida arat dtherm
features: 029ee3ff-bfebfbff-00000001-2c100800
features_pv: 17c9cbf5-82b82203-2191cbf5-00000003-00000000-00000000-00000000-00000000-00001000-8c000000
features_hvm: 17cbfbff-82ba2223-2d93fbff-00000003-00000000-00000000-00000000-00000000-00001000-8c000000

 

How could I know exactly where the problem comes from? Because you can see how the features_hvm has a variation from 9c000000 to 8c000000. At BIOS level I use default values activating the virtualization.

On the other hand reviewing more in depth differences the file "/proc/cpuinfo" I have observed that in the master of the pool marks the following:

bugs: l1tf

 

While in the new host it does not appear:

bugs:

 

Could this be the problem? Do you know any possibility to fix it to the current pool without affecting the availability? Because it is a pool in production.

 

Thank you very much Alan and Tobias for your answers!

[Tobias Kreidl]

Is perhaps the BIOS level newer on one machine than the other?

 

As to the bugs entry, there may be a hardware issue; see for example: https://unix.stackexchange.com/questions/456425/what-does-the-bugs-section-of-proc-cpuinfo-actually-show

 

-=Tobias

[Alan Lantz]

l1tf is a CPU microcode issue. You might have to contact your hardware vendor for either microcode to

fix the issue or a replacement CPUs. I'm guessing it would be your older CPU's that have the bug that

has been addressed by the newere CPU's you are trying to add. 

 

https://access.redhat.com/articles/3562741

 

[Sergey Polyakov]

Try to check your virtualization settings in BIOS

I had a similar problem due to incorrect power settings (which disabled virtualization)