seitan 的档案 - activity

Hello, I'm having difficulties to find how to configure openstack-ansible to make all cinder services use glance_api_version = 2
By default openstack-ansible configures cinder with
glance_api_version = 1
Thank you for your help.

Just to update: the patch for this bug was accepted and committed to Ocata: https://review.openstack.org/#/c/472266/

For anyone, that will come across similar problem, there is more explanation about it (and ways to fix it) here:

https://bugs.launchpad.net/nova/+bug/1687581 (https://bugs.launchpad.net/nova/+bug/...)

I must add, that this happens on all compute nodes, - there are no nova errors. Nova just announces that Instance destroyed successfully. instead of starting it, followed by oslo_messaging.rpc.server errors.

Hello, after upgrading OpenStack from Newton to Ocata via openstack-ansible, I cannot start any of existing servers:

openstack server list |grep a6200dff-2010-44a2-92ca-3af303ec3577

| a6200dff-2010-44a2-92ca-3af303ec3577 | Win10VDI-3-ephemeral | SHUTOFF | internal=10.0.0.17|

openstack server start a6200dff-2010-44a2-92ca-3af303ec3577
openstack server list |grep a6200dff-2010-44a2-92ca-3af303ec3577

| a6200dff-2010-44a2-92ca-3af303ec3577 | Win10VDI-3-ephemeral | SHUTOFF | internal=10.0.0.17|

Server is still in SHUTOFF state.
nova.log contains the flowing entries:

http://paste.ubuntu.com/24465172/

What could cause this? Thank you.

It seems to be related to bug #1677168:
https://bugs.launchpad.net/charm-nova-cloud-controller/+bug/1677168 (https://bugs.launchpad.net/charm-nova...)

nova-manage cell_v2 list_cells
nova-manage cell_v2 map_instances --cell_uuid <uuid for cell1>

In nova-api-metadata container fixes problem above.

Hello, after successfully upgrading openstack from newton to ocata by using steps, described in:
https://docs.openstack.org/developer/openstack-ansible/ocata/upgrade-guide/manual-upgrade.html (https://docs.openstack.org/developer/...)

I have problem starting servers:

openstack server list |grep SHUTOFF

| a6200dff-2010-44a2-92ca-3af303ec3577 | Win10VDI-3-ephemeral | SHUTOFF | internal=10.0.0.17  |



openstack server start a6200dff-2010-44a2-92ca-3af303ec3577 
No server with a name or ID of 'a6200dff-2010-44a2-92ca-3af303ec3577' exists.


nova start  a6200dff-2010-44a2-92ca-3af303ec3577 
No server with a name or ID of 'a6200dff-2010-44a2-92ca-3af303ec3577' exists.
ERROR (CommandError): Unable to start the specified server(s).

Where should I look into at first? Thank you.

Hello, I'm trying to run tox tests on nova. Every test passes fine, but pip-missing-reqs fails:

tox -e pip-missing-reqs

Traceback (most recent call last):
  File "/root/nova/.tox/pip-missing-reqs/bin/pip-missing-reqs", line 7, in <module>
    from pip_missing_reqs.find_missing_reqs import main
  File "/root/nova/.tox/pip-missing-reqs/local/lib/python2.7/site-packages/pip_missing_reqs/find_missing_reqs.py", line 14, in <module>
    from pip.utils import get_installed_distributions, normalize_name
ImportError: cannot import name normalize_name
ERROR: InvocationError: '/root/nova/.tox/pip-missing-reqs/bin/pip-missing-reqs -d --ignore-file=nova/tests/* --ignore-file=nova/test.py nova'
___________________________________________________________________________________________________________________________________ summary ___________________________________________________________________________________________________________________________________
ERROR:   pip-missing-reqs: commands failed

pip --version pip 9.0.1 from /usr/local/lib/python2.7/dist-packages (python 2.7)

This happens on Debian Jessie, Ubuntu Trusty and Ubuntu Xenial. Any ideas, what's wrong? Thank you.

Hello,
I'm trying to understand the console management via OpenStack REST API. Documentation is very cheap on this. Lets say, I want to list all consoles for server:

Array
(
[url] => http://192.168.0.2:8774/v2.1/87f8c0ed08534b1da737f85d055a0912/servers/a17ac994-8311-42ab-84d0-614e1ef8f1cd/consoles
[content_type] => application/json
[http_code] => 200
[header_size] => 308
[request_size] => 346
[filetime] => -1
[ssl_verify_result] => 0
[redirect_count] => 0
[total_time] => 0.028263
[namelookup_time] => 5.3E-5
[connect_time] => 0.000552
[pretransfer_time] => 0.000609
[size_upload] => 0
[size_download] => 16
[speed_download] => 566
[speed_upload] => 0
[download_content_length] => 16
[upload_content_length] => -1
[starttransfer_time] => 0.028232
[redirect_time] => 0
[redirect_url] =>
[primary_ip] => 192.168.0.2
[certinfo] => Array
(
) ` [primary_port] => 8774` ` [local_ip] => 192.168.0.13` ` [local_port] => 51416` ` [request_header] => GET /v2.1/87f8c0ed08534b1da737f85d055a0912/servers/a17ac994-8311-42ab-84d0-614e1ef8f1cd/consoles HTTP/1.1` `Host: 192.168.0.2:8774` `Accept: */*` `X-Auth-Token: gAAAAABY0PIjJnRRH6alc8BFVY8j4xjBekjeUG58Chq-081BLZ-2ydiYw5t9vtH_7f8lzys2vECpa8moghsedatma18CCUrq6Z6Hqat1iAuccAJWKqVOOn91NRiz6ZJtHlUx_PEpIi84xhk28Trh08C09cU3kttoPP9FVpRzNwiuzBZEG7VuA0s`
)

The reply I get is empty console array:

{"consoles": []}

Though vm instance is accessible via SPICE-hml5 console. So is this normal?

Also, what does "Create Console" call do? If I post empty request to /servers/{server_id}/consoles I get no error, but i cant see any visible changes to virtual machine. Hope anyone can help me here. Thank you.

Since it appears, that OpenStack uses its own library to generate XML config for virtual machine and there was no way to override emulator path, a new blueprint was added to Launchpad:

https://blueprints.launchpad.net/nova/+spec/custom-emulator-path (https://blueprints.launchpad.net/nova...)

Hello, I would like to change emulator for newly created machines from current /usr/bin/qemu-system-x86_64 to wrapper script /usr/bin/somescript. Where are default emulator paths defined? Thank you

According to: https://docs.hpcloud.com/hos-3.x/helion/releasenotes30.html

This is not an error and should be ignored.

Source: https://bugs.launchpad.net/openstack-ansible/+bug/1653483

I've updated system via openstack-ansible 14.0.5, yet problem persists, all nova-compute nodes spam logs with 'rbd python libraries not found' error.

Hello, I'm deploying openstack via openstack-ansible. Install went okay, - I've got a working openstack cluster. After that i've decided to make some changes - after rerunning playbooks, I'm getting errors in nova compute nodes:

2017-01-09 09:15:37.668 23007 ERROR nova.compute.manager [req-4dbdf38f-2786-48cd-a866-18e7f85ac88b - - - - -] No compute node record for host ostack-ibm2
2017-01-09 09:15:37.671 23007 INFO nova.compute.resource_tracker [req-4dbdf38f-2786-48cd-a866-18e7f85ac88b - - - - -] Auditing locally available compute resources for node ostack-ibm2.mgmt.domain.com
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager [req-4dbdf38f-2786-48cd-a866-18e7f85ac88b - - - - -] Error updating resources for node ostack-ibm2.mgmt.domain.com.
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager Traceback (most recent call last):
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/compute/manager.py", line 6415, in update_available_resource_for_node
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     rt.update_available_resource(context)
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/compute/resource_tracker.py", line 511, in update_available_resource
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     resources = self.driver.get_available_resource(self.nodename)
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 5280, in get_available_resource
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     disk_info_dict = self._get_local_gb_info()
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 4916, in _get_local_gb_info
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     info = LibvirtDriver._get_rbd_driver().get_pool_info()
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/virt/libvirt/driver.py", line 994, in _get_rbd_driver
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     rbd_user=CONF.libvirt.rbd_user)
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager   File "/openstack/venvs/nova-14.0.3/lib/python2.7/site-packages/nova/virt/libvirt/storage/rbd_utils.py", line 130, in __init__
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager     raise RuntimeError(_('rbd python libraries not found'))
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager RuntimeError: rbd python libraries not found
2017-01-09 09:15:37.672 23007 ERROR nova.compute.manager

I'm using openstack-ansible 14.0.3. Has anyone had the same problem? Thank you.

Hello, I'm deploying my openstack system via openstack-ansible.
Everything works fine, but it seems that /etc/keystone/ssl/certs/ in keystone containers is empty.
I'm using rados to provide object storage to openstack.
According to http://docs.ceph.com/docs/jewel/rados...
you need to generate certificates from /etc/keystone/ssl/certs/ for rados to be able to decode keystone messages.

I'm guessing because of that, rados fails to query keystone with the following errors:

2016-12-23 12:01:30.972648 7f5aaf7d6700  0 revoked tokens response is missing signed section
2016-12-23 12:01:30.972724 7f5aaf7d6700  0 ERROR: keystone revocation processing returned error r=-22

Is there any additional parameter I should use to in my ansible configuration to make those certificates be created?

my rados ceph.conf:

[client.rgw.ostack-rados1.domain.com]
rgw_frontends = "civetweb port=80"
rgw keystone url = http://192.168.0.1:5000
rgw keystone admin user = ceph
rgw keystone admin password = secret
rgw keystone admin project = admin
rgw keystone admin domain = default
rgw keystone api version = 3
rgw keystone token cache size = 500
rgw keystone revocation interval = 500
rgw s3 auth use keystone = true
rgw keystone verify ssl = false

Thank you for your help.

Thank you. Managed to get object storage up and running. Sharing a small howto on this:

http://www.neblogas.lt/2016/12/23/openstack-object-storage-ceph/ (http://www.neblogas.lt/2016/12/23/ope...)

Thank you for reply. So I was doing this the wrong way then. I just have to configure radosGW and add it into keystone then?

Hello, I've installed OpenStack via openstack-ansible. Everything woks fine. I'm using ceph storage as secondary storage for VMs. I wonder, is there a playbook to install swift object storage and configure it to use ceph as storage backend? Thank you.