stackhpc/tenks

Virtual bare metal cluster management

Tenks

Tenks is a utility that manages virtual bare metal clusters for development and testing purposes. It offers:

• Declarative definition of clusters using Ansible's familiar YAML syntax
• Automated creation and deletion of virtualised bare metal nodes across any number of hypervisors
• In-built management of Open vSwitch bridges for physical networks
• Registration of nodes and ports with OpenStack Ironic, and creation of OpenStack Nova flavors
• Simple cluster reconfiguration and tear-down

To get started with Tenks, see the documentation (linked below).

• Free software: Apache license
• Documentation: https://tenks.readthedocs.io/
• Source: https://github.com/stackhpc/tenks/
• Bugs: https://github.com/stackhpc/tenks/issues
• Known issues/limitations: refer to the documentation.

tenks's Issues

teardown fails if node is in clean-failed state

TASK [ironic-enrolment : Configure node in Ironic] ************************************************************************************************************
fatal: [localhost]: FAILED! => {"changed": false, "msg": "Error unregistering node 'befce008-31ad-488f-a3bf-3e129ac97875' due to current provision state 'clean failed'"}

Should we put it into the manage state before hand?

Failure setting up openvswitch

On centos 7.5 running tenks in an all in one install with a kayobe deployed openstack cloud, I'm seeing:

TASK [Ensure Open vSwitch is started and enabled] ******************************
[DEPRECATION WARNING]: state=running is deprecated. Please use state=started. 
This feature will be removed in version 2.7. Deprecation warnings can be 
disabled by setting deprecation_warnings=False in ansible.cfg.
fatal: [localhost]: FAILED! => {"changed": false, "msg": "Unable to enable service openvswitch: Failed to execute operation: Cannot send after transport endpoint shutdown\n"}

This is occurred after a reboot.

Some things I tried:

(tenks) 2 centos@kayobe-will:~/tenks$ sudo systemctl restart openvswitch       
Failed to restart openvswitch.service: Unit is masked.

(tenks) 2 centos@kayobe-will:~/tenks$ sudo systemctl unmask openvswitch       

which didn't seem to help.

disabling the service allowed me to get past this task:

sudo systemctl disable openvswitch

but the service doesn't look healthy:

sudo systemctl status openvswitch -l
● openvswitch.service - Open vSwitch
   Loaded: loaded (/usr/lib/systemd/system/openvswitch.service; disabled; vendor preset: disabled)
   Active: inactive (dead)

Fails at configure node in ironic

I'm seeing:

TASK [ironic-enrolment : Configure node in Ironic] *****************************
fatal: [localhost]: FAILED! => {"changed": false, "msg": "Baremetal machine node failed to be created.: Client Error for url: http://192.168.122.210/baremetal/v1/nodes, Forbidden"}

This is on a ubuntu machine which devstack deployed.

stack@ubuntu:~/tenks$ uname -a
Linux ubuntu 4.4.0-116-generic #140-Ubuntu SMP Mon Feb 12 21:23:04 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux

stack@ubuntu:~/devstack$ git log -n 1
commit 991b1f13f0aed578dccec2e761be69005357444f
Author: melanie witt <[email protected]>
Date:   Thu Sep 27 18:30:38 2018 +0000

    Update cinder backup_driver to full class name
    
    Legacy backup service support was recently dropped from cinder in
    change I3ada2dee1857074746b1893b82dd5f6641c6e579 and we need to
    adjust how we set the config option in devstack accordingly. This
    updates the backup_driver option to specify a full class name instead
    of only the module name.
    
    Closes-Bug: #1794859
    
    Change-Id: I3a72f38b564b8b83b233fccba7685833b6394d45

Ansible version:

stack@ubuntu:~/devstack$ ansible --version
ansible 2.6.4
  config file = None
  configured module search path = [u'/opt/stack/.ansible/plugins/modules', u'/usr/share/ansible/plugins/modules']
  ansible python module location = /opt/stack/tenks/.venv/local/lib/python2.7/site-packages/ansible
  executable location = /opt/stack/tenks/.venv/bin/ansible
  python version = 2.7.12 (default, Dec  4 2017, 14:50:18) [GCC 5.4.0 20160609]

openstack variables in the environment:

OS_PROJECT_DOMAIN_ID=default
OS_REGION_NAME=RegionOne
OS_USER_DOMAIN_ID=default
OS_PROJECT_NAME=demo
OS_IDENTITY_API_VERSION=3
OS_PASSWORD=password
OS_AUTH_TYPE=password
OS_AUTH_URL=http://192.168.122.210/identity
OS_USERNAME=demo
OS_TENANT_NAME=demo
OS_VOLUME_API_VERSION=3
OS_CACERT=
``