Open Stack

Check status of nova-compute at,

# openstack compute service list

If the compute node is UP, check for nova-compute logs at compute node for exact error.

Hope this helps.

/
  Trinath Somanchi,
   NXP SDN/NFV Team.

From: huanglingyan (A) [mailto:huanglingyan2 at huawei.com]
Sent: Tuesday, December 19, 2017 2:18 PM
To: openstack at lists.openstack.org
Subject: [Openstack] devstack：update libvirt to 3.4.0

hi all,

hoping someone could shed some light on this issue I’ve worked on several days.

I have installed openstack through devstack and it works well when I create an instance.

stack at ubuntu:/var/run$ openstack server list
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+
| ID                                   | Name  | Status | Networks                       | Image                    | Flavor  |
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+
| 6d6ac274-0f0b-4576-8f54-07e006214c9b | ins_1 | ACTIVE | public=172.24.4.4, 2001:db8::3 | cirros-0.3.5-x86_64-disk | m1.tiny |
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+


The default version of libvirt is 2.5.0 while I want to use libvirt 3.4.0. What I have done is to install libvirt 3.4.0 manully since no apt repo has such version of libvirt.

To check libvirt version,

virsh # version
Compiled against library: libvirt 3.4.0
Using library: libvirt 3.4.0
Using API: QEMU 3.4.0
Running hypervisor: QEMU 2.9.0

However, my openstack seems to lose connetion with the new version of libvirt. It shows error when I create a new instance.

stack at ubuntu:/var/run$ openstack server list
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+
| ID                                   | Name  | Status | Networks                       | Image                    | Flavor  |
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+
| 59369bec-7015-449b-bd4e-85f3237a3228 | ins_2 | ERROR  |                                | cirros-0.3.5-x86_64-disk | m1.tiny |
| 6d6ac274-0f0b-4576-8f54-07e006214c9b | ins_1 | ACTIVE | public=172.24.4.4, 2001:db8::3 | cirros-0.3.5-x86_64-disk | m1.tiny |
+--------------------------------------+-------+--------+--------------------------------+--------------------------+---------+

The log shows

Failed to schedule instances:

Traceback (most recent call last):
  File "/usr/local/lib/python2.7/dist-packages/oslo_messaging/rpc/server.py", line 232, in inner
    return func(*args, **kwargs)
  File "/home/stack/nova/nova/scheduler/manager.py", line 149, in select_destinations
    alloc_reqs_by_rp_uuid, provider_summaries)
  File "/home/stack/nova/nova/scheduler/filter_scheduler.py", line 110, in select_destinations
    raise exception.NoValidHost(reason=reason)

NoValidHost: No valid host was found. There are not enough hosts available.

ERROR nova.conductor.manager Traceback (most recent call last):
ERROR nova.conductor.manager   File "/home/stack/nova/nova/conductor/manager.py", line 1034, in schedule_and_build_ins
ERROR nova.conductor.manager     instance_uuids)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/conductor/manager.py", line 626, in _schedule_instances
ERROR nova.conductor.manager     request_spec, instance_uuids)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/scheduler/utils.py", line 586, in wrapped
ERROR nova.conductor.manager     return func(*args, **kwargs)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/scheduler/client/__init__.py", line 52, in select_destinati
ERROR nova.conductor.manager     instance_uuids)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/scheduler/client/__init__.py", line 37, in __run_method
ERROR nova.conductor.manager     return getattr(self.instance, __name)(*args, **kwargs)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/scheduler/client/query.py", line 33, in select_destinations
ERROR nova.conductor.manager     instance_uuids)
ERROR nova.conductor.manager   File "/home/stack/nova/nova/scheduler/rpcapi.py", line 137, in select_destinations
ERROR nova.conductor.manager     return cctxt.call(ctxt, 'select_destinations', **msg_args)
ERROR nova.conductor.manager   File "/usr/local/lib/python2.7/dist-packages/oslo_messaging/rpc/client.py", line 169, i
ERROR nova.conductor.manager     retry=self.retry)
ERROR nova.conductor.manager   File "/usr/local/lib/python2.7/dist-packages/oslo_messaging/transport.py", line 123, in
ERROR nova.conductor.manager     timeout=timeout, retry=retry)
ERROR nova.conductor.manager   File "/usr/local/lib/python2.7/dist-packages/oslo_messaging/_drivers/amqpdriver.py", li
ERROR nova.conductor.manager     retry=retry)
ERROR nova.conductor.manager   File "/usr/local/lib/python2.7/dist-packages/oslo_messaging/_drivers/amqpdriver.py", li
ERROR nova.conductor.manager     raise result
ERROR nova.conductor.manager NoValidHost_Remote: No valid host was found. There are not enough hosts available.


What can I do to let openstack connect to the new libvirt?
thanks!!


--huanglingyan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack/attachments/20171219/a26e9012/attachment.html>