Bug 1233023

Summary: Guest can not be started when vcpu placement auto is set
Product: Red Hat Enterprise Linux 7 Reporter: Dan Zheng <dzheng>
Component: libvirtAssignee: Andrea Bolognani <abologna>
Status: CLOSED NOTABUG QA Contact: Virtualization Bugs <virt-bugs>
Severity: medium Docs Contact:
Priority: high    
Version: 7.2CC: dyuan, dzheng, gsun, lhuang, mzhan, rbalakri, zhwang
Target Milestone: rc   
Target Release: ---   
Hardware: ppc64le   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2015-08-14 06:17:53 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:

Description Dan Zheng 2015-06-18 05:33:10 UTC 
Description of problem:
When vcpu placement auto is set, guest can not be started.

Version-Release number of selected component (if applicable):
libvirt-1.2.16-1.el7.ppc64le
qemu-kvm-rhev-2.3.0-2.el7.ppc64le
kernel-3.10.0-229.ael7b.ppc64le

How reproducible:
100%

Steps to Reproduce:
1. Prepare a guest with XMl setting below:
 <vcpu placement='auto'>4</vcpu>

2. Start the guest and get failure
# virsh start dzhengvm1
error: Failed to start domain dzhengvm1
error: invalid argument: Failed to parse bitmap ''

3. Check libvirtd.log
 ---libvirtd.log---
2015-06-18 01:45:39.714+0000: 88235: error : virCommandWait:2552 : internal error: Child process (/bin/numad -w 4:1000) unexpected exit status 1
2015-06-18 01:45:39.714+0000: 88235: error : virNumaGetAutoPlacementAdvice:72 : internal error: Failed to query numad for the advisory nodeset
2015-06-18 01:45:39.714+0000: 88235: error : virBitmapParse:423 : invalid argument: Failed to parse bitmap ''

4. Check numad can run standalone and works correctly. 
# numad -w 4:1000
0

5. Check guest log
2015-06-18 05:22:00.396+0000: shutting down

Actual results:
Guest fails to start with below errors:
error: Failed to start domain dzhengvm1
error: invalid argument: Failed to parse bitmap ''



Expected results:
The guest should be able to start.

Additional info:
[root@ibm-p8-kvm-02-qe /]# numactl --hardware
available: 4 nodes (0-1,16-17)
node 0 cpus: 0 8 16 24 32
node 0 size: 65536 MB
node 0 free: 45897 MB
node 1 cpus: 40 48 56 64 72
node 1 size: 65536 MB
node 1 free: 48205 MB
node 16 cpus: 80 88 96 104 112
node 16 size: 65536 MB
node 16 free: 47374 MB
node 17 cpus: 120 128 136 144 152
node 17 size: 65536 MB
node 17 free: 53089 MB
node distances:
node   0   1  16  17 
  0:  10  20  40  40 
  1:  20  10  40  40 
 16:  40  40  10  20 
 17:  40  40  20  10 

[root@ibm-p8-kvm-02-qe /]# virsh nodeinfo
CPU model:           ppc64le
CPU(s):              20
CPU frequency:       3690 MHz
CPU socket(s):       1
Core(s) per socket:  160
Thread(s) per core:  1
NUMA cell(s):        1
Memory size:         263635264 KiB
Comment 2 Wayne Sun 2015-06-30 08:48:26 UTC 
This is not 100% reproducible

Tested with
pkgs:
libvirt-1.2.16-1.el7.ppc64le
qemu-kvm-rhev-2.3.0-5.el7.ppc64le
kernel-3.10.0-282.el7.ppc64le

# virsh dumpxml virt-tests-debug-clone
...
  <vcpu placement='auto'>5</vcpu>
...

# virsh start virt-tests-debug-clone
Domain virt-tests-debug-clone started
Comment 3 Andrea Bolognani 2015-08-13 09:48:24 UTC 
After a dozen attempts I wasn't able to reproduce the issue.

I'm using the following packages:

  kernel-3.10.0-302.el7.ppc64le
  qemu-kvm-rhev-2.3.0-16.el7.ppc64le
  libvirt-daemon-1.2.17-4.el7.ppc64le

Do you still run into the issue on an updated host?
Comment 4 Dan Zheng 2015-08-14 02:01:02 UTC 
(In reply to Andrea Bolognani from comment #3)
> After a dozen attempts I wasn't able to reproduce the issue.
> 
> I'm using the following packages:
> 
>   kernel-3.10.0-302.el7.ppc64le
>   qemu-kvm-rhev-2.3.0-16.el7.ppc64le
>   libvirt-daemon-1.2.17-4.el7.ppc64le
> 
> Do you still run into the issue on an updated host?

Using above packages, I can not reproduce it any more. Pls close it. Thanks.