Red Hat Bugzilla – Bug 1096196
QEMU should abort if NUMA node configuration don't cover all RAM
Last modified: 2015-03-05 04:46:01 EST
Description of problem: ---------------------------- NUMA emulation failed with extra memory not specified in any node. Version-Release number of selected component (if applicable): ---------------------------------------------------------------- qemu-kvm-1.5.3-60.el7_0.1.x86_64 How reproducible: ------------------ 100% Steps to Reproduce: ----------------------- boot guest with memory lager than specified in all nodes, like: -m 5G -numa node,mem=1024,cpus=0,nodeid=0 -numa node,mem=1024,cpus=1,nodeid=1 -numa node,mem=1024,cpus=2,nodeid=2 -numa node,mem=1024,cpus=3,nodeid=3 Actual results: ------------------ in guest: [root@localhost ~]# numactl --hardware available: 1 nodes (0) node 0 cpus: 0 1 2 3 node 0 size: 5119 MB node 0 free: 4525 MB node distances: node 0 0: 10 (qemu) info numa 4 nodes node 0 cpus: 0 node 0 size: 1024 MB node 1 cpus: 1 node 1 size: 1024 MB node 2 cpus: 2 node 2 size: 1024 MB node 3 cpus: 3 node 3 size: 1024 MB With "-m 4G", everything is fine. I recall long time ago(we do not have a case cover the scenario) i tested this kind of scenario, if there is extra memory not specified in any node, it resides in node 0 by default, and other node is as specified, i am not sure whether it should work like this now, if this is as designed, feel free to close it.
Not a bug, QEMU is doing exactly what was requested by the user. But the guest chooses to ignore the NUMA topology because it doesn't make sense. There are plans to abort QEMU in case the user provides such invalid configuration, though, so I am keeping the BZ open to track the new feature.
Moving to qemu-kvm-rhev. Patches were included upstream and will be on QEMU 2.1.0.
reproduce this bug with unfix version: qemu-kvm-rhev-1.5.3-60.el7_0.10.x86_64 use command : /usr/libexec/qemu-kvm -m 5G -numa node,mem=1024,cpus=0,nodeid=0 -numa node,mem=1024,cpus=1,nodeid=1 -numa node,mem=1024,cpus=2,nodeid=2 -numa node,mem=1024,cpus=3,nodeid=3 result: guest do not abort. (qemu) info numa 4 nodes node 0 cpus: 0 node 0 size: 1024 MB node 1 cpus: node 1 size: 1024 MB node 2 cpus: node 2 size: 1024 MB node 3 cpus: node 3 size: 1024 MB [root@localhost ~]# numactl --hardware available: 4 nodes (0-3) node 0 cpus: 0 2 4 6 8 10 12 14 node 0 size: 8157 MB node 0 free: 2572 MB node 1 cpus: 16 18 20 22 24 26 28 30 node 1 size: 8192 MB node 1 free: 2036 MB node 2 cpus: 1 3 5 7 9 11 13 15 node 2 size: 8192 MB node 2 free: 2086 MB node 3 cpus: 17 19 21 23 25 27 29 31 node 3 size: 8175 MB node 3 free: 3173 MB node distances: node 0 1 2 3 0: 10 16 16 16 1: 16 10 16 16 2: 16 16 10 16 3: 16 16 16 10 so, this bug is reproduced. verify with version : qemu-kvm-rhev-2.1.2-5.el7.x86_64 steps: 1.execute command: /usr/libexec/qemu-kvm -m 5G -numa node,mem=1024,cpus=0,nodeid=0 -numa node,mem=1024,cpus=1,nodeid=1 -numa node,mem=1024,cpus=2,nodeid=2 -numa node,mem=1024,cpus=3,nodeid=3 result: qemu abort and outputs the following info: qemu-kvm: total memory for NUMA nodes (0x100000000) should equal RAM size (0x140000000) according to above test result. this bug is fixed.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://rhn.redhat.com/errata/RHSA-2015-0624.html