Bug 1126777
| Summary: | guest which set numa in xml can't start success | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Luyao Huang <lhuang> |
| Component: | qemu-kvm-rhev | Assignee: | Eduardo Habkost <ehabkost> |
| Status: | CLOSED ERRATA | QA Contact: | Virtualization Bugs <virt-bugs> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | 7.1 | CC: | dyuan, hhuang, huding, jiahu, jmiao, juzhang, knoel, lhuang, mprivozn, mrezanin, mzhan, rbalakri, virt-maint, xfu |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | x86_64 | ||
| OS: | All | ||
| Whiteboard: | |||
| Fixed In Version: | QEMU 2.1.0 | Doc Type: | Bug Fix |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2015-03-05 09:48:51 UTC | Type: | Bug |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
I don't think this is a libvirt bug. The qemu process was started successfully. The only problem I can see from libvirt POV is that the sum of memory defined under <numa/> exceeds the overall memory defined unde <memory/>. But if that's a problem for qemu, it should refuse to start. Starting on QEMU 2.1, it will refuse to start when an invalid NUMA config like the above is provided. In either case, QEMU is simply doing exactly what you asked for. Some guests will just complain about the meaningless NUMA config you provided, some guests may crash. Reproduce bug with qemu-kvm-1.5.3-69.el7.x86_64 Boot guest with invalid numa configuration(NUMA nodes memory isn't equal RAM size) ... -m 2048 -realtime mlock=off -smp 4,sockets=4,cores=1,threads=1 -numa node,nodeid=0,cpus=0-1,mem=509 -numa node,nodeid=1,cpus=2-3,mem=509 ... result: guest can start normally. Verify this bug with qemu-kvm-rhev-2.1.0-2.el7.x86_64 Tested two scenarios. 1.Boot guest with invalid numa configuration(NUMA nodes memory don't equal RAM size) -m 2048 -realtime mlock=off -smp 4,sockets=4,cores=1,threads=1 -numa node,nodeid=0,cpus=0-1,mem=509 -numa node,nodeid=1,cpus=2-3,mem=509 Result: Guest can not start. and qemu-kvm show warning message. qemu-kvm: total memory for NUMA nodes (1067450368) should equal RAM size (80000000) 2.Boot guest with valid numa configuration(NUMA nodes memory equal RAM size) -m 1018 -realtime mlock=off -smp 4,sockets=4,cores=1,threads=1 -numa node,nodeid=0,cpus=0-1,mem=509 -numa node,nodeid=1,cpus=2-3,mem=509 Result: Guest start successfully. Another, QE tested it with virt-manager as well, and got the same result as above. 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 |
Description of problem: guest which set numa in xml can't start success when set memory='101200' Version-Release number of selected component (if applicable): libvirt-1.2.7-1.el7.x86_64 kernel-3.10.0-140.el7.x86_64 How reproducible: 100% Steps to Reproduce: 1.prepare a guest with : # virsh dumpxml test6 <memory unit='KiB'>1042432</memory> <currentMemory unit='KiB'>1041857</currentMemory> <vcpu placement='static'>4</vcpu> ........... <cpu> <numa> <cell cpus='0-1' memory='1012000'/> <cell cpus='2-3' memory='1012000'/> </numa> </cpu> ........ 2.# virsh start test6 Domain test6 started 3.Use virt-manager check guest status and guest will cannot start normal. error in guest: PANIC：early exception 06 rip 10:ffffffffff81c4adb6 error 0 cr2 0 Actual results: Guest OS can't start success and there only a error message in guest: PANIC：early exception 06 rip 10:ffffffffff81c4adb6 error 0 cr2 0 Expected results: Guest OS can start success Additional info: when set memory=512000 or memory=1312000 ,guest can start normal.