Remove virtualbmc specific environment
By default tripleo-quickstart creates virtual environments managed by VirtualBMC. The documentation for manually creating VirtualBMC environments predates common usage of quickstart, so now it is redundant. This change removes the virtualbmc environment, and refers to the virtual environment instead, where appropriate. Change-Id: If4fe81444a70f147ccd4239247025cd5b68c744c
This commit is contained in:
Steve Baker
parent
7c4ff6f931
commit
f1a77d1ede
@ -321,23 +321,8 @@ Generic Hardware Types
|
||||
.. admonition:: Virtual
|
||||
:class: virtual
|
||||
|
||||
When combined with :doc:`virtualbmc`, this hardware type can be used for
|
||||
developing and testing TripleO in a virtual environment as well.
|
||||
|
||||
.. admonition:: Stable Branch
|
||||
:class: stable
|
||||
|
||||
Prior to the Ocata release, a special ``pxe_ssh`` driver was used for
|
||||
testing Ironic in the virtual environment. This driver connects to the
|
||||
hypervisor to conduct management operations on virtual nodes. In case
|
||||
of this driver, ``pm_addr`` is a hypervisor address, ``pm_user`` is
|
||||
a SSH user name for accessing hypervisor, ``pm_password`` is a private
|
||||
SSH key for accessing hypervisor. Note that private key must not be
|
||||
encrypted.
|
||||
|
||||
.. warning::
|
||||
The ``pxe_ssh`` driver is deprecated and ``pxe_ipmitool`` +
|
||||
:doc:`virtualbmc` should be used instead.
|
||||
This hardware type can be used for developing and testing TripleO in a
|
||||
:doc:`virtual` as well.
|
||||
|
||||
* Another generic hardware type is redfish_. It provides support for the
|
||||
quite new `Redfish standard`_, which aims to replace IPMI eventually as
|
||||
|
||||
@ -10,4 +10,3 @@ section contains instructions on how to setup your environments properly.
|
||||
standalone
|
||||
virtual
|
||||
baremetal
|
||||
virtualbmc
|
||||
|
||||
@ -1,142 +0,0 @@
|
||||
VirtualBMC
|
||||
==========
|
||||
|
||||
VirtualBMC_ is a small CLI that allows users to create a virtual BMC_
|
||||
to manage a virtual machines using the IPMI_ protocol, similar to how real
|
||||
bare metal machines are managed. It can be used to to enable testing
|
||||
bare metal deployments in completely virtual environments.
|
||||
|
||||
.. admonition:: Stable Branch
|
||||
:class: stable
|
||||
|
||||
Ironic also ships a ``pxe_ssh`` driver that can be used for that purpose,
|
||||
but it has been deprecated and its use is discouraged.
|
||||
|
||||
.. warning::
|
||||
VirtualBMC is not meant for production environments.
|
||||
|
||||
Installation
|
||||
------------
|
||||
|
||||
VirtualBMC is available from RDO repositories starting with the Ocata release::
|
||||
|
||||
sudo yum install -y python-virtualbmc
|
||||
|
||||
It is usually installed and used on the hypervisor where the virtual machines
|
||||
reside.
|
||||
|
||||
.. _create-vbmc:
|
||||
|
||||
Creating virtual BMC
|
||||
--------------------
|
||||
|
||||
Every virtual machine needs its own virtual BMC. Create it with::
|
||||
|
||||
vbmc add <domain> --port 6230 --username admin --password password
|
||||
|
||||
.. note::
|
||||
You need to use a different port for each domain. Port numbers
|
||||
lower than 1025 requires the user to have root privilege in the system.
|
||||
|
||||
.. note::
|
||||
For **tripleo-quickstart** you may have to specify
|
||||
``--libvirt-uri=qemu:///session``.
|
||||
|
||||
Start the virtual BMCs::
|
||||
|
||||
vbmc start <domain>
|
||||
|
||||
.. warning::
|
||||
This step has to be repeated after virtual host reboot.
|
||||
|
||||
Test the virtual BMC to see if it's working. For example, to power on
|
||||
the virtual machine do::
|
||||
|
||||
ipmitool -I lanplus -U admin -P password -H 127.0.0.1 -p 6230 power on
|
||||
|
||||
Enrolling virtual machines
|
||||
--------------------------
|
||||
|
||||
In the undercloud, populate the ``instackenv.json`` with new "bare metals"
|
||||
in a similar way to real bare metal machines (see :ref:`instackenv`) with
|
||||
two exceptions:
|
||||
|
||||
* set ``pm_port`` to the port you specified when `Creating virtual BMC`
|
||||
|
||||
* populate ``mac`` field even if you plan to use introspection
|
||||
|
||||
For example:
|
||||
|
||||
.. code-block:: json
|
||||
|
||||
{
|
||||
"nodes": [
|
||||
{
|
||||
"pm_type": "ipmi",
|
||||
"mac": [
|
||||
"00:0a:f2:88:12:aa"
|
||||
],
|
||||
"pm_user": "admin",
|
||||
"pm_password": "password",
|
||||
"pm_addr": "172.16.0.1",
|
||||
"pm_port": "6230",
|
||||
"name": "compute-0"
|
||||
}
|
||||
]
|
||||
}
|
||||
|
||||
Migrating from pxe_ssh to VirtualBMC
|
||||
------------------------------------
|
||||
|
||||
If you already have a virtual cloud deployed and want to migrate from the
|
||||
deprecated ``pxe_ssh`` driver to ``ipmi`` using VirtualBMC,
|
||||
follow `Creating virtual BMC`_, then update the existing nodes to change
|
||||
their drivers and certain driver properties:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
openstack baremetal node set $NODE_UUID_OR_NAME \
|
||||
--driver ipmi \
|
||||
--driver-info ipmi_address=<IP address of the virthost> \
|
||||
--driver-info ipmi_port=<Virtual BMC port> \
|
||||
--driver-info ipmi_username="admin" \
|
||||
--driver-info ipmi_password="password"
|
||||
|
||||
.. admonition:: Stable Branch
|
||||
:class: stable
|
||||
|
||||
For the Ocata release, use ``pxe_ipmitool`` driver instead of ``ipmi``.
|
||||
|
||||
In the case of bare metal service in the overcloud, you will first have to
|
||||
configure the deployment to include the pxe_ipmitool driver, then rerun the
|
||||
deployment command,
|
||||
for example:
|
||||
|
||||
.. code-block:: yaml
|
||||
|
||||
parameter_defaults:
|
||||
IronicEnabledDrivers:
|
||||
- pxe_ipmitool
|
||||
- pxe_ssh
|
||||
|
||||
|
||||
Before updating to Pike release, make sure to remove the pxe_ssh driver from the
|
||||
deployment configuration, as it will be removed from Ironic, then rerun
|
||||
the deployment command,
|
||||
for example:
|
||||
|
||||
.. code-block:: yaml
|
||||
|
||||
parameter_defaults:
|
||||
IronicEnabledDrivers:
|
||||
- pxe_ipmitool
|
||||
|
||||
To validate after updating deployment and verify everything is populated properly:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
openstack baremetal node validate $NODE_UUID_OR_NAME | grep power
|
||||
|
||||
.. _VirtualBMC: https://opendev.org/openstack/virtualbmc
|
||||
.. _IPMI: https://en.wikipedia.org/wiki/Intelligent_Platform_Management_Interface
|
||||
.. _BMC: https://en.wikipedia.org/wiki/Baseboard_management_controller
|
||||
@ -55,12 +55,6 @@ TripleO workflows. E.g. for node using IPMI::
|
||||
TripleO-specific, e.g. they force local boot instead of network boot used
|
||||
by default in Ironic.
|
||||
|
||||
.. note::
|
||||
If you use :doc:`../environments/virtualbmc`, make sure to follow
|
||||
:ref:`create-vbmc` for the overcloud nodes as well, and correctly populate
|
||||
``ipmi_port``. If needed, change ``ipmi_address`` to the address of the
|
||||
virtual host, which is accessible from controllers.
|
||||
|
||||
Then enroll only ``undercloud.json`` in your undercloud::
|
||||
|
||||
source stackrc
|
||||
|
||||
@ -46,7 +46,7 @@ registration process (which may hang due to this), it means that Ironic is
|
||||
unable to verify power management credentials, and you need to fix them.
|
||||
Check the ``pm_addr``, ``pm_user`` and ``pm_password`` fields in your
|
||||
``instackenv.json``. In some cases (e.g. when using
|
||||
:doc:`../environments/virtualbmc`) you also need a correct ``pm_port``.
|
||||
:doc:`../environments/virtual`) you also need a correct ``pm_port``.
|
||||
Update the node as explained in `Fixing invalid node information`_.
|
||||
|
||||
Fixing invalid node information
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user