Created attachment 1105535 [details] engine log Description of problem: Failed to run vm with numa pinning on host with one numa node under strict mode Version-Release number of selected component (if applicable): rhevm-3.6.1.3-0.1.el6.noarch How reproducible: Always Steps to Reproduce: 1. Create vm(1024GB, 1 cpu) and pin it to host with one numa node 2. Create one numa node on vm and pin it to host numa node 3. Start vm Actual results: Vm failed to start Error while executing action: test_numa: Cannot run VM. There is no host that satisfies current scheduling constraints. See below for details: The host cyan-vdsg.qa.lab.tlv.redhat.com did not satisfy internal filter Memory because cannot accommodate memory of VM's pinned virtual NUMA nodes within host's physical NUMA nodes.. Expected results: vm succeed to run Additional info: vm numa nodes via REST <vm_numa_nodes> <vm_numa_node href="/ovirt-engine/api/vms/c7ecd2dc-dbd3-4419-956f-1249651c0f2b/numanodes/4a86ac4f-e08b-494c-9c23-78f5ff9e102b" id="4a86ac4f-e08b-494c-9c23-78f5ff9e102b"> <index>0</index> <memory>1024</memory> <cpu> <cores> <core index="0" /> </cores> </cpu> <vm href="/ovirt-engine/api/vms/c7ecd2dc-dbd3-4419-956f-1249651c0f2b" id="c7ecd2dc-dbd3-4419-956f-1249651c0f2b" /> <numa_node_pins> <numa_node_pin pinned="true" index="0"> <host_numa_node id="417cdefb-8c47-4838-87f3-dd0498fdf6c7" /> </numa_node_pin> </numa_node_pins> </vm_numa_node> </vm_numa_nodes> numa from host: # numactl -H available: 1 nodes (0) node 0 cpus: 0 1 node 0 size: 4094 MB node 0 free: 3287 MB node distances: node 0 0: 10
I am pretty sure I ran into that problem also with more than one numa node. Will check it again. From the UI perspective the problem will be fixed with https://bugzilla.redhat.com/show_bug.cgi?id=1289967 but the canDoAction should prevent the pinning too when there is only one host.
Please add a can-do-action check for no NUMA / single node.
Meital, can you test this when you get a chance? It's been neglected for a long time, but Roman's comment indicates that the UX may already be fixed
deprecating SLA team usage, moving to Virt
We didn't get to this bug for more than 2 years, and it's not being considered for the upcoming 4.4. It's unlikely that it will ever be addressed so I'm suggesting to close it. If you feel this needs to be addressed and want to work on it please remove cond nack and target accordingly.
In addition I was not able to simulate this issue.
Closing old bug. Please reopen if still relevant/you want to work on it.
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 1000 days