Document URL: 
https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/7/html/virtualization_tuning_and_optimization_guide/sect-virtualization_tuning_optimization_guide-numa-numa_and_libvirt#sect-Virtualization_Tuning_Optimization_Guide-NUMA-Guest_NUMA_multinode-hugepages

Section Number and Name: 
9.3.10. Assigning Host Huge Pages to Multiple Guest NUMA Nodes

Describe the issue: 
Maybe I am wrong but I think this is incorrect. Assigning Host Huge Pages to Multiple Guest NUMA Nodes is only supported with 'qemu-kvm-rhev' package provided by OSP or RHV subscriptions, not with 'qemu-kvm' package shipped with standard RHEL subscriptions.

If you try a similar setup just as described in section 9.3.10 you will get the following error when trying to start the virtual machine:
~~~
error: Failed to start domain rhel74
error: unsupported configuration: huge pages per NUMA node are not supported with this QEMU
~~~  

Setting up guest NUMA topology as described in section 9.3.9 and using huge pages are supported when used separately, but not when using them together as stated in section 9.3.10.

Suggestions for improvement:


Additional information:
 
For the host huge pages configuration, I followed the steps in section "Procedure 8.3. Allocating 1 GB huge pages at runtime" in the same document.

For the vm the configurations used for this test were:
~~~
...
   <numa>
      <cell id='0' cpus='0-1' memory='2048000' unit='KiB'/>
      <cell id='1' cpus='2-3' memory='2048000' unit='KiB'/>
    </numa>
...

  <memoryBacking>
    <hugepages>
      <page size='1048576' unit='KiB' nodeset='0'/>
      <page size='2048' unit='KiB' nodeset='1'/>
    </hugepages>
  </memoryBacking>
...
~~~ 

And
~~~
...
   <numa>
      <cell id='0' cpus='0-1' memory='2048000' unit='KiB'/>
      <cell id='1' cpus='2-3' memory='2048000' unit='KiB'/>
    </numa>
...

  <memoryBacking>
    <hugepages/>
  </memoryBacking>
...
~~~

Both failed with the same error:
~~~
error: Failed to start domain rhel74
error: unsupported configuration: huge pages per NUMA node are not supported with this QEMU
~~~

When trying to use guest numa topology or huge pages separately with the corresponding configuration fragment it works.