Question

Failed to Power on guest VMs running on CE 5.6 (ISO installer method)


Userlevel 1
Badge +8
Just deployed CE 5.6 using the new ISO installer (ce-2018.05.01-installer.iso) but failed to power on the guest VMs.

Error Msg:

Operation failed: InternalException: UncaughtException: Traceback (most recent call last): File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/util/task/base_task.py", line 525, in _resume File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/hypervisor/host.py", line 74, in wrapper File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 313, in _run File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 598, in _power_on File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 514, in _power_state_transition_helper File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 469, in _power_state_transition_iterative_helper File "/usr/lib64/python2.7/contextlib.py", line 24, in __exit__ self.gen.next() File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 450, in _power_state_transition_iterative_helper File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/vm/set_power_state_task.py", line 562, in _do_power_on File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/hypervisor/kvm/libvirt_connection.py", line 151, in wrapper File "/home/afg/src/main/builds/build-ce-2018.05.01-stable-release/python-tree/bdist.linux-x86_64/egg/acropolis/hypervisor/kvm/power_mixin.py", line 186, in power_on_vm File "build/bdist.linux-x86_64/egg/pylibvirt.py", line 425, in createXML cimpl = self._pool.call("DomainCreateXML", self._cimpl, xml_bytes, flags) File "build/bdist.linux-x86_64/egg/pylibvirt.py", line 111, in call raise libvirtError(*err) libvirtError: Failed to create controller cpu for group: No such file or directory

This topic has been closed for comments

22 replies

Userlevel 2
Badge +7
Same here, everything works great except user VMs.
But I installed CE without a dedicated boot drive for AHV, now reinstalling on USB sticks.
Userlevel 2
Badge +7
Reinstall didn't help, still getting "libvirtError: Failed to create controller cpu for group: No such file or directory".
Userlevel 2
Badge +7
Solved by image-based installation - it's much faster and actually working.
Userlevel 2
Badge +7
I am facing this issue as well, I would love to have a fix, as the ISO method it is much faster and versatile than the image one, please help us @aluciani
Userlevel 2
Badge +7
Doing more troubleshooting, I can see the next, which might be related:
code:
[root@NTNX-34309415-A ~]# libvirtd -v
2018-06-18 20:06:39.279+0000: 3144: info : libvirt version: 2.0.0, package: 12.20180425.1.23736.el7 (Unknown, 2018-04-26-01:47:48, 9d316c537538)
2018-06-18 20:06:39.279+0000: 3144: info : hostname: NTNX-34309415-A
2018-06-18 20:06:39.279+0000: 3144: info : virObjectNew:202 : OBJECT_NEW: obj=0x5576eae06360 classname=virAccessManagerClass
2018-06-18 20:06:39.279+0000: 3144: info : virObjectNew:202 : OBJECT_NEW: obj=0x5576eadfc600 classname=virAccessManagerClass
2018-06-18 20:06:39.279+0000: 3144: info : virObjectRef:296 : OBJECT_REF: obj=0x5576eae06360
2018-06-18 20:06:39.279+0000: 3144: info : virObjectUnref:259 : OBJECT_UNREF: obj=0x5576eae06360
2018-06-18 20:06:39.279+0000: 3144: error : virPidFileAcquirePath:422 : Failed to acquire pid file '/var/run/libvirtd.pid': Resource temporarily unavailable
2018-06-18 20:06:39.279+0000: 3144: info : virNetlinkEventServiceStopAll:715 : stopping all netlink event services
Userlevel 2
Badge +7
I think the ISO has multiple issues that needs to be addressed, first:
This come as N:
code:
cat /sys/module/kvm_intel/parameters/nested
N

So you need to run the next in order to enable it:
code:
echo "options kvm-intel nested=1" | sudo tee /etc/modprobe.d/kvm-intel.conf


The below error on my previous post couldn't be fixed on my side. Also, when I try to list the VMs, my new VM it is not on the list, just the CVM, but I have another one on my PRISM:
code:
virsh # list --all
Id Name State
----------------------------------------------------
1 NTNX-34309415-A-CVM running
Userlevel 2
Badge +7
Another update, I did a yum update on the Acropolis, which shouldn't be very traumatic, but in this case it did upgraded from libvirtd 2.0.0 to the latest 3.9.0 the error remains
code:
libvirtd -v
2018-06-18 23:18:14.676+0000: 1677: info : libvirt version: 3.9.0, package: 14.el7_5.5 (CentOS BuildSystem , 2018-05-22-02:42:56, c1bm.rdu2.centos.org)
2018-06-18 23:18:14.676+0000: 1677: info : hostname: NTNX-76d76d6c-A
2018-06-18 23:18:14.676+0000: 1677: info : virObjectNew:254 : OBJECT_NEW: obj=0x55c75e77f000 classname=virAccessManagerClass
2018-06-18 23:18:14.677+0000: 1677: info : virObjectNew:254 : OBJECT_NEW: obj=0x55c75e772950 classname=virAccessManagerClass
2018-06-18 23:18:14.677+0000: 1677: info : virObjectRef:388 : OBJECT_REF: obj=0x55c75e77f000
2018-06-18 23:18:14.677+0000: 1677: info : virObjectUnref:350 : OBJECT_UNREF: obj=0x55c75e77f000
2018-06-18 23:18:14.677+0000: 1677: error : virPidFileAcquirePath:422 : Failed to acquire pid file '/var/run/libvirtd.pid': Resource temporarily unavailable
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStopAll:781 : stopping all netlink event services
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStop:744 : stopping netlink event service
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStop:744 : stopping netlink event service
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStop:744 : stopping netlink event service
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStop:744 : stopping netlink event service
2018-06-18 23:18:14.677+0000: 1677: info : virNetlinkEventServiceStop:744 : stopping netlink event service


And not just this, the CVM now it seems that is not working properly so I can not connect to it.
Userlevel 2
Badge +7
It might be partitioning/permissions issue. On three node cluster with 6 SSDs and 3 USB sticks in total I'm getting all 6 SSDs with dd available for ADFS, but only 4 with ISO installation method.
Userlevel 2
Badge +7
Maybe @jpiscaer have some insights? No one is trying CE 5.6 over vSphere? @PiPoe2H ?
Badge +4
I have it running but with the traditional way. With the ISO installer I have the same issue.

I did some troubleshooting and I think this is not related to the machine_type or PMU, since I forced to 7.2 and same issue. Also, if you use virsh to create a VM it works with no issue.
Userlevel 2
Badge +11
Maybe @jpiscaer have some insights? No one is trying CE 5.6 over vSphere? @PiPoe2H ?
No real insights, other than that I've had contact with the developer and he's aware of a few known bugs relating to partitioning in the ISO install method. My best guess is there will be a fix for these soon. If they're related to the inability to power on VMs, I do not know.
Userlevel 2
Badge +7
Any update @aluciani ? 🙂
Userlevel 2
Badge +7
Any update @aluciani ?
Userlevel 3
Badge +4
Hitting the same issue ("Failed to create controller cpu for group: No such file or directory") with 2018.05.01 ISO installer nested on Nutanix AHV.
Is there any easy fix for this without having to rebuild the CE clusters?

Kind regards,
Lukas
Userlevel 6
Badge +16
I think ISO installation is more in test version, try regular install if you really need CE.
Userlevel 2
Badge +7
Any update @aluciani ? 🙂 🙂
Userlevel 7
Badge +34
Looking into it at the moment - hope to hear something soon.
Userlevel 2
Badge +11
I had the same issue. Rebuilt a node using the ISO then found I couldn't migrate to or start any VM's on it. After reviewing this discussion, tried the image install which worked fine.

I did like the ISO install procedure... I look forward to it's issues being resolved.
Badge +4
Hi everyone... is there anybody who has solved the problem in the meantime? I still can't poweron any vm... 😞
Userlevel 6
Badge +16
Check this thread similar issues with powering on VM:

https://next.nutanix.com/discussion-forum-14/vm-is-not-booting-with-attached-iso-in-nutanix-community-edition-28513#post32490
Userlevel 6
Badge +16
If above thread doesn't help you I would suggest you to install CE with regular image installation.
Badge +4
I used a regular installation... I'm continuing to write here

https://next.nutanix.com/discussion-forum-14/error-power-on-a-vm-aos-5-6-ce-28464/index1.html#post32545