Minor docs changes for better readability and consistency
Change-Id: I43a281231da6f254528c52961805ceef23358aa1
This commit is contained in:
Mahnoor Asghar
Riccardo Pittau
parent
a94f595c94
commit
bcd694e90c
31
README.rst
31
README.rst
@ -2,25 +2,24 @@
|
||||
Redfish development tools
|
||||
=========================
|
||||
|
||||
This is a set of simple simulation tools aiming at supporting the
|
||||
development and testing of the Redfish protocol implementations and,
|
||||
in particular, Sushy library (https://docs.openstack.org/sushy/). It
|
||||
is not designed for use outside of development and testing environments.
|
||||
Please do not run sushy-tools in a production environment of any kind.
|
||||
This is a set of simple simulation tools aimed at supporting the development and
|
||||
testing of the Redfish protocol implementations and, in particular, the Sushy
|
||||
library (https://docs.openstack.org/sushy/). It is not designed for use outside
|
||||
of development and testing environments. Please do not run sushy-tools in a
|
||||
production environment of any kind.
|
||||
|
||||
The package ships two simulators - static Redfish responder and
|
||||
virtual Redfish BMC that is backed by libvirt or OpenStack cloud.
|
||||
The package ships two simulators - the static Redfish responder and the virtual
|
||||
Redfish BMC (which is backed by libvirt or OpenStack cloud).
|
||||
|
||||
The static Redfish responder is a simple REST API server which
|
||||
responds the same things to client queries. It is effectively
|
||||
read-only.
|
||||
The static Redfish responder is a simple REST API server which responds with the
|
||||
same things to client queries. It is effectively read-only.
|
||||
|
||||
The virtual Redfish BMC resembles the real Redfish-controlled bare-metal
|
||||
machine to some extent. Some client queries are translated to commands that
|
||||
actually control VM instances simulating bare metal hardware. However some
|
||||
of the Redfish commands just return static content never touching the
|
||||
virtualization backend and, for that matter, virtual Redfish BMC is similar
|
||||
to the static Redfish responder.
|
||||
The virtual Redfish BMC resembles the real Redfish-controlled bare metal machine
|
||||
to some extent. Some client queries are translated to commands that actually
|
||||
control VM instances simulating bare metal hardware. However, some of the
|
||||
Redfish commands just return static content, never touching the virtualization
|
||||
backend and in this regard, the virtual Redfish BMC is similar to the static
|
||||
Redfish responder.
|
||||
|
||||
* Free software: Apache license
|
||||
* Documentation: https://docs.openstack.org/sushy-tools
|
||||
|
||||
@ -1,4 +1,4 @@
|
||||
# sushy emulator configuration file build on top of Flask application
|
||||
# sushy emulator configuration file built on top of Flask application
|
||||
# configuration infrastructure: http://flask.pocoo.org/docs/config/
|
||||
|
||||
# Listen on all local IP interfaces
|
||||
@ -25,18 +25,16 @@ SUSHY_EMULATOR_IRONIC_CLOUD = None
|
||||
# The libvirt URI to use. This option enables libvirt driver.
|
||||
SUSHY_EMULATOR_LIBVIRT_URI = u'qemu:///system'
|
||||
|
||||
# Instruct the libvirt driver to ignore any instructions to
|
||||
# set the boot device. Allowing the UEFI firmware to instead
|
||||
# rely on the EFI Boot Manager
|
||||
# Note: This sets the legacy boot element to dev="fd"
|
||||
# and relies on the floppy not existing, it likely won't work
|
||||
# your VM has a floppy drive.
|
||||
# Instruct the libvirt driver to ignore any instructions to set the boot device,
|
||||
# allowing the UEFI firmware to instead rely on the EFI Boot Manager.
|
||||
# Note: This sets the legacy boot element to dev="fd" and relies on the floppy
|
||||
# not existing. It likely won't work if your VM has a floppy drive.
|
||||
SUSHY_EMULATOR_IGNORE_BOOT_DEVICE = False
|
||||
|
||||
|
||||
# The map of firmware loaders dependent on the boot mode and
|
||||
# system architecture. Ideally the x86_64 loader will be capable
|
||||
# of secure boot or not based on the chosen nvram.
|
||||
# The map of firmware loaders dependent on the boot mode and system
|
||||
# architecture. Ideally the x86_64 loader will be capable of secure boot or not
|
||||
# based on the chosen nvram.
|
||||
SUSHY_EMULATOR_BOOT_LOADER_MAP = {
|
||||
'UEFI': {
|
||||
'x86_64': u'/usr/share/OVMF/OVMF_CODE.secboot.fd',
|
||||
@ -52,14 +50,13 @@ SUSHY_EMULATOR_BOOT_LOADER_MAP = {
|
||||
SUSHY_EMULATOR_SECURE_BOOT_ENABLED_NVRAM = '/usr/share/OVMF/OVMF_VARS.secboot.fd'
|
||||
SUSHY_EMULATOR_SECURE_BOOT_DISABLED_NVRAM = '/usr/share/OVMF/OVMF_VARS.fd'
|
||||
|
||||
# This map contains statically configured Redfish Chassis linked
|
||||
# up with the Systems and Managers enclosed into this Chassis.
|
||||
# This map contains statically configured Redfish Chassis linked up with the
|
||||
# Systems and Managers enclosed into this Chassis.
|
||||
#
|
||||
# The first chassis in the list will contain all other resources.
|
||||
#
|
||||
# If this map is not present in the configuration, a single default
|
||||
# Chassis is configured automatically to enclose all available Systems
|
||||
# and Managers.
|
||||
# If this map is not present in the configuration, a single default Chassis is
|
||||
# configured automatically to enclose all available Systems and Managers.
|
||||
SUSHY_EMULATOR_CHASSIS = [
|
||||
{
|
||||
u'Id': u'Chassis',
|
||||
@ -68,23 +65,22 @@ SUSHY_EMULATOR_CHASSIS = [
|
||||
}
|
||||
]
|
||||
|
||||
# This map contains statically configured Redfish IndicatorLED
|
||||
# resource state ('Lit', 'Off', 'Blinking') keyed by UUIDs of
|
||||
# System and Chassis resources.
|
||||
# This map contains statically configured Redfish IndicatorLED resource state
|
||||
# ('Lit', 'Off', 'Blinking'), keyed by UUIDs of System and Chassis resources.
|
||||
#
|
||||
# If this map is not present in the configuration, each
|
||||
# System and Chassis will have their IndicatorLED `Lit` by default.
|
||||
# If this map is not present in the configuration, each System and Chassis will
|
||||
# have their IndicatorLED `Lit` by default.
|
||||
#
|
||||
# Redfish client can change IndicatorLED state. The new state
|
||||
# is volatile, i.e. it's maintained in process memory.
|
||||
# The Redfish client can change IndicatorLED state. The new state is volatile,
|
||||
# i.e. it's maintained in process memory.
|
||||
SUSHY_EMULATOR_INDICATOR_LEDS = {
|
||||
# u'48295861-2522-3561-6729-621118518810': u'Blinking'
|
||||
}
|
||||
|
||||
# This map contains statically configured virtual media resources.
|
||||
# These devices ('Cd', 'Floppy', 'USBStick') will be exposed by the
|
||||
# Manager(s) and possibly used by the System(s) if system emulation
|
||||
# backend supports boot image configuration.
|
||||
# These devices ('Cd', 'Floppy', 'USBStick') will be exposed by the Manager(s)
|
||||
# and possibly used by the System(s) if system emulation backend supports boot
|
||||
# image configuration.
|
||||
#
|
||||
# This value is ignored by the OpenStack driver, which only supports the 'Cd'
|
||||
# device. If this map is not present in the configuration, the following
|
||||
@ -106,12 +102,12 @@ SUSHY_EMULATOR_VMEDIA_DEVICES = {
|
||||
}
|
||||
}
|
||||
|
||||
# Instruct the virtual media insertion not to verify the SSL certificate
|
||||
# when retrieving the image.
|
||||
# Instruct the virtual media insertion to not verify the SSL certificate when
|
||||
# retrieving the image.
|
||||
SUSHY_EMULATOR_VMEDIA_VERIFY_SSL = False
|
||||
|
||||
# This map contains statically configured Redfish Storage resource linked
|
||||
# up with the Systems resource, keyed by the UUIDs of the Systems.
|
||||
# This map contains statically configured Redfish Storage resources linked up
|
||||
# with the Systems resources, keyed by the UUIDs of the Systems.
|
||||
SUSHY_EMULATOR_STORAGE = {
|
||||
"da69abcc-dae0-4913-9a7b-d344043097c0": [
|
||||
{
|
||||
@ -131,10 +127,10 @@ SUSHY_EMULATOR_STORAGE = {
|
||||
]
|
||||
}
|
||||
|
||||
# This map contains statically configured Redfish Drives resource. The Drive
|
||||
# This map contains statically configured Redfish Drives resources. The Drive
|
||||
# objects are keyed in a composite fashion using a tuple of the form
|
||||
# (System_UUID, Storage_ID) referring to the UUID of the System and Id of the
|
||||
# Storage resource, respectively, to which the drive belongs.
|
||||
# Storage resource, respectively, to which the Drive belongs.
|
||||
SUSHY_EMULATOR_DRIVES = {
|
||||
("da69abcc-dae0-4913-9a7b-d344043097c0", "1"): [
|
||||
{
|
||||
@ -146,20 +142,20 @@ SUSHY_EMULATOR_DRIVES = {
|
||||
]
|
||||
}
|
||||
|
||||
# This map contains dynamically configured Redfish Volume resource backed
|
||||
# by the libvirt virtualization backend of the dynamic Redfish emulator.
|
||||
# The Volume objects are keyed in a composite fashion using a tuple of the
|
||||
# form (System_UUID, Storage_ID) referring to the UUID of the System and ID
|
||||
# of the Storage resource, respectively, to which the Volume belongs.
|
||||
# This map contains dynamically configured Redfish Volume resources backed by
|
||||
# the libvirt virtualization backend of the dynamic Redfish emulator.
|
||||
# The Volume objects are keyed in a composite fashion using a tuple of the form
|
||||
# (System_UUID, Storage_ID) referring to the UUID of the System and ID of the
|
||||
# Storage resource, respectively, to which the Volume belongs.
|
||||
#
|
||||
# Only the volumes specified in the map or created via a POST request are
|
||||
# Only the Volumes specified in the map or created via a POST request are
|
||||
# allowed to be emulated upon by the emulator. Volumes other than these can
|
||||
# neither be listed nor deleted.
|
||||
#
|
||||
# The Volumes from map missing in the libvirt backend will be created
|
||||
# The Volumes in the map missing from the libvirt backend will be created
|
||||
# dynamically in the pool name specified (provided the pool exists in the
|
||||
# backend). If the pool name is not specified, the volume will be created
|
||||
# automatically in pool named 'default'.
|
||||
# backend). If the pool name is not specified, the Volume will be created
|
||||
# automatically in a pool named 'default'.
|
||||
SUSHY_EMULATOR_VOLUMES = {
|
||||
('da69abcc-dae0-4913-9a7b-d344043097c0', '1'): [
|
||||
{
|
||||
@ -183,7 +179,7 @@ SUSHY_EMULATOR_VOLUMES = {
|
||||
|
||||
# This list contains the identities of instances that the driver will filter by.
|
||||
# It is useful in a tenant environment where only some instances represent
|
||||
# virtual baremetal.
|
||||
# virtual bare metal.
|
||||
SUSHY_EMULATOR_ALLOWED_INSTANCES = [
|
||||
"437XR1138R2",
|
||||
"1",
|
||||
@ -195,4 +191,4 @@ SUSHY_EMULATOR_ALLOWED_INSTANCES = [
|
||||
|
||||
# Disable the ability to power off the node, in line with NCSI enablement in
|
||||
# Ironic
|
||||
SUSHY_EMULATOR_DISABLE_POWER_OFF = False
|
||||
SUSHY_EMULATOR_DISABLE_POWER_OFF = False
|
||||
|
||||
@ -2,15 +2,14 @@
|
||||
Configuring emulators
|
||||
=====================
|
||||
|
||||
Running emulators in background
|
||||
-------------------------------
|
||||
Running emulators in the background
|
||||
-----------------------------------
|
||||
|
||||
The emulators run as interactive processes attached to the
|
||||
terminal by default. You can create a systemd service to run the
|
||||
emulators in background.
|
||||
For each emulator create a systemd unit file and
|
||||
update full path to ``sushy-static`` or ``sushy-emulator`` binary and
|
||||
adjust arguments as necessary, for example::
|
||||
The emulators run as interactive processes attached to the terminal by default.
|
||||
You can create systemd services to run the emulators in the background.
|
||||
For each emulator, create a systemd unit file, and update ``<full-path>`` to the
|
||||
``sushy-static`` or ``sushy-emulator`` binary, and adjust the arguments as
|
||||
necessary, for example::
|
||||
|
||||
[Unit]
|
||||
Description=Sushy Libvirt emulator
|
||||
@ -22,9 +21,9 @@ adjust arguments as necessary, for example::
|
||||
StandardOutput=syslog
|
||||
StandardError=syslog
|
||||
|
||||
If you want to run emulators with different configuration, for example,
|
||||
``sushy-static`` emulator with different mockup files, then create a new
|
||||
systemd unit file.
|
||||
If you want to run the emulators with different configurations, for example, the
|
||||
``sushy-static`` emulator with different mockup files, then create a new systemd
|
||||
unit file.
|
||||
|
||||
You can also use ``gunicorn`` to run ``sushy-emulator``, for example::
|
||||
|
||||
@ -34,15 +33,15 @@ Using configuration file
|
||||
------------------------
|
||||
|
||||
Besides command-line options, `sushy-emulator` can be configured via a
|
||||
configuration file. The tool uses Flask application
|
||||
`configuration infrastructure <http://flask.pocoo.org/docs/config/>`_,
|
||||
emulator-specific configuration options are prefixed with **SUSHY_EMULATOR_**
|
||||
to make sure they won't collide with Flask's own configuration options.
|
||||
configuration file. This tool uses the Flask application
|
||||
`configuration infrastructure <http://flask.pocoo.org/docs/config/>`_.
|
||||
Emulator-specific configuration options are prefixed with **SUSHY_EMULATOR_**
|
||||
to make sure that they don't collide with Flask's own configuration options.
|
||||
|
||||
The configuration file itself can be specified through the
|
||||
`SUSHY_EMULATOR_CONFIG` environment variable.
|
||||
|
||||
The full list of supported options and their meanings could be found in
|
||||
the sample configuration file:
|
||||
The full list of supported options and their meanings can be found in the sample
|
||||
configuration file:
|
||||
|
||||
.. literalinclude:: emulator.conf
|
||||
|
||||
@ -9,7 +9,7 @@ Contributing
|
||||
Cloning the sushy-tools repository
|
||||
++++++++++++++++++++++++++++++++++
|
||||
|
||||
If you haven't already, sushy-tools source code should be pulled directly
|
||||
If you haven't already, the sushy-tools source code should be pulled directly
|
||||
from git.
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -16,17 +16,17 @@ Or, if you have virtualenvwrapper installed:
|
||||
$ mkvirtualenv sushy-tools
|
||||
$ pip install sushy-tools
|
||||
|
||||
The *Virtual Redfish BMC* tool relies upon one or more hypervisors to mimic
|
||||
bare metal nodes. Depending on the virtualization backend you are planning
|
||||
to use, certain third-party dependencies should also be installed.
|
||||
The *Virtual Redfish BMC* tool relies upon one or more hypervisors to mimic bare
|
||||
metal nodes. Depending on the virtualization backend you are planning to use,
|
||||
certain third-party dependencies should also be installed.
|
||||
|
||||
The dependencies for the virtualization backends that should be installed
|
||||
for the corresponding drivers to become operational are:
|
||||
The dependencies for the virtualization backends that should be installed for
|
||||
the corresponding drivers to become operational are:
|
||||
|
||||
* `libvirt-python` for the libvirt driver
|
||||
* `openstacksdk` for the nova driver
|
||||
|
||||
.. note::
|
||||
|
||||
The dependencies for at least one virtualization backend should be
|
||||
satisfied to have *Virtual Redfish BMC* emulator operational.
|
||||
The dependencies for at least one virtualization backend should be satisfied
|
||||
to have the *Virtual Redfish BMC* emulator operational.
|
||||
|
||||
@ -2,11 +2,10 @@
|
||||
Virtual Redfish BMC
|
||||
===================
|
||||
|
||||
The Virtual Redfish BMC is functionally similar to the
|
||||
`Virtual BMC <https://opendev.org/openstack/virtualbmc>`_ tool
|
||||
except that the frontend protocol is Redfish rather than IPMI. The Redfish
|
||||
commands coming from the client are handled by one or more resource-specific
|
||||
drivers.
|
||||
The Virtual Redfish BMC emulator is functionally similar to the
|
||||
`Virtual BMC <https://opendev.org/openstack/virtualbmc>`_ tool, except that the
|
||||
frontend protocol is Redfish rather than IPMI. The Redfish commands coming from
|
||||
the client are handled by one or more resource-specific drivers.
|
||||
|
||||
Feature sets
|
||||
------------
|
||||
@ -24,20 +23,19 @@ Supported feature sets are:
|
||||
Systems resource
|
||||
----------------
|
||||
|
||||
For *Systems* resource, emulator maintains two drivers relying on
|
||||
a virtualization backend to emulate bare metal machines by means of
|
||||
virtual machines. In addition, there is a fake driver used to mock
|
||||
bare metal machines.
|
||||
For the *Systems* resource, the emulator maintains two drivers relying on a
|
||||
virtualization backend to emulate bare metal machines by means of virtual
|
||||
machines. In addition, there is a fake driver used to mock bare metal machines.
|
||||
|
||||
The following sections will explain how to configure and use
|
||||
each of these drivers.
|
||||
The following sections will explain how to configure and use each of these
|
||||
drivers.
|
||||
|
||||
Systems resource driver: libvirt
|
||||
++++++++++++++++++++++++++++++++
|
||||
|
||||
First thing you need is to set up some libvirt-managed virtual machines
|
||||
(AKA domains) to manipulate. The following command will create a new
|
||||
virtual machine i.e. libvirt domain `vbmc-node`:
|
||||
The first thing you need is to set up some libvirt-managed virtual machines (AKA
|
||||
domains) to manipulate. The following command will create a new virtual machine
|
||||
i.e. libvirt domain `vbmc-node`:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -62,8 +60,7 @@ Next you can fire up the Redfish virtual BMC which will listen at
|
||||
sushy-emulator
|
||||
* Running on http://localhost:8000/ (Press CTRL+C to quit)
|
||||
|
||||
Now you should be able to see your libvirt domain among the Redfish
|
||||
*Systems*:
|
||||
Now you should be able to see your libvirt domain among the Redfish *Systems*:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -158,15 +155,15 @@ Redfish Simple Storage Controllers available for this domain):
|
||||
UEFI boot
|
||||
~~~~~~~~~
|
||||
|
||||
By default, `legacy` or `BIOS` mode is used to boot the instance. However,
|
||||
By default, `legacy` or `BIOS` mode is used to boot the instance. However, the
|
||||
libvirt domain can be configured to boot via UEFI firmware. This process
|
||||
requires additional preparation on the host side.
|
||||
|
||||
On the host you need to have OVMF firmware binaries installed. Fedora users
|
||||
could pull them as `edk2-ovmf` RPM. On Ubuntu, `apt-get install ovmf` should
|
||||
do the job.
|
||||
could pull them as `edk2-ovmf` RPM. On Ubuntu, `apt-get install ovmf` should do
|
||||
the job.
|
||||
|
||||
Then you need to create a VM by running `virt-install` with the UEFI specific
|
||||
Then you need to create a VM by running `virt-install` with the UEFI-specific
|
||||
`--boot` options:
|
||||
|
||||
Example:
|
||||
@ -191,8 +188,9 @@ Example:
|
||||
virsh define --file $tmpfile
|
||||
rm $tmpfile
|
||||
|
||||
This will create a new `libvirt` domain with path to OVMF images properly
|
||||
configured. Let's take a note on the path to the blob by running `virsh dumpxml vbmc-node`:
|
||||
This will create a new `libvirt` domain with the path to OVMF images properly
|
||||
configured. Let's take a note on the path to the blob by running
|
||||
`virsh dumpxml vbmc-node`:
|
||||
|
||||
Example:
|
||||
|
||||
@ -209,10 +207,10 @@ Example:
|
||||
...
|
||||
</domain>
|
||||
|
||||
Because now we need to add this path to emulator's configuration matching VM
|
||||
architecture we are running. It is also possible to make Redfish calls to enable
|
||||
or disable Secure Boot by specifying which nvram template to load in each case.
|
||||
Make a copy of stock configuration file and edit it accordingly:
|
||||
Because now we need to add this path to the emulator's configuration matching
|
||||
the VM architecture we are running. It is also possible to make Redfish calls to
|
||||
enable or disable Secure Boot by specifying which nvram template to load in each
|
||||
case. Make a copy of the stock configuration file and edit it accordingly:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -240,9 +238,8 @@ Now you can run `sushy-emulator` with the updated configuration file:
|
||||
Settable boot image
|
||||
~~~~~~~~~~~~~~~~~~~
|
||||
|
||||
The `libvirt` system emulation backend supports setting custom boot images,
|
||||
so that libvirt domains (representing bare metal nodes) can boot from user
|
||||
images.
|
||||
The `libvirt` system emulation backend supports setting custom boot images, so
|
||||
that libvirt domains (representing bare metal nodes) can boot from user images.
|
||||
|
||||
This feature enables system boot from virtual media device.
|
||||
|
||||
@ -256,21 +253,21 @@ virtual media boot.
|
||||
Systems resource driver: OpenStack
|
||||
++++++++++++++++++++++++++++++++++
|
||||
|
||||
You can use an OpenStack cloud instances to simulate Redfish-managed
|
||||
baremetal machines. This setup is known under the name of
|
||||
You can use OpenStack cloud instances to simulate Redfish-managed bare metal
|
||||
machines. This setup is known under the name of
|
||||
`OpenStack Virtual Baremetal <http://openstack-virtual-baremetal.readthedocs.io/en/latest/>`_.
|
||||
We will largely reuse its OpenStack infrastructure and configuration
|
||||
instructions. After all, what we are trying to do here is to set up the
|
||||
Redfish emulator alongside the
|
||||
instructions. After all, what we are trying to do here is to set up the Redfish
|
||||
emulator alongside the
|
||||
`openstackbmc <https://github.com/cybertron/openstack-virtual-baremetal/blob/master/openstack_virtual_baremetal/openstackbmc.py>`_
|
||||
tool which is used for exactly the same purpose at OVB with the only
|
||||
difference that it works over the *IPMI* protocol as opposed to *Redfish*.
|
||||
tool which is used for exactly the same purpose at OVB with the only difference
|
||||
being that it works over the *IPMI* protocol as opposed to *Redfish*.
|
||||
|
||||
The easiest way is probably to set up your OpenStack Virtual Baremetal cloud
|
||||
by following
|
||||
The easiest way is probably to set up your OpenStack Virtual Baremetal cloud by
|
||||
following
|
||||
`its instructions <http://openstack-virtual-baremetal.readthedocs.io/en/latest/>`_.
|
||||
|
||||
Once your OVB cloud operational, you log into the *BMC* instance and
|
||||
Once your OVB cloud is operational, you log into the *BMC* instance and
|
||||
:ref:`set up sushy-tools <installation>` there.
|
||||
|
||||
Next you can invoke the Redfish virtual BMC pointing it to your OVB cloud:
|
||||
@ -302,7 +299,8 @@ Redfish *Systems*:
|
||||
"@Redfish.Copyright": "Copyright 2014-2016 Distributed Management Task Force, Inc. (DMTF). For the full DMTF copyright policy, see http://www.dmtf.org/about/policies/copyright."
|
||||
}
|
||||
|
||||
And flip its power state via the Redfish call:
|
||||
|
||||
And flip an instance's power state via the Redfish call:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -321,19 +319,19 @@ Systems resource driver: Ironic
|
||||
++++++++++++++++++++++++++++++++++
|
||||
|
||||
You can use the Ironic driver to manage Ironic baremetal instance to simulated
|
||||
Redfish API. You may want to do this if you require a redfish compatible endpoint
|
||||
but don't have direct access to the BMC (you only have access via Ironic) or
|
||||
the BMC doesn't support redfish.
|
||||
Redfish API. You may want to do this if you require a redfish-compatible
|
||||
endpoint but don't have direct access to the BMC (you only have access via
|
||||
Ironic) or the BMC doesn't support redfish.
|
||||
|
||||
Assuming your baremetal cloud is setup you can invoke the Redfish emulator by
|
||||
running
|
||||
Assuming your bare metal cloud is set up you can invoke the Redfish emulator by
|
||||
running:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
sushy-emulator --ironic-cloud baremetal-cloud
|
||||
* Running on http://localhost:8000/ (Press CTRL+C to quit)
|
||||
|
||||
By this point you should be able to see your Baremetal instances among the
|
||||
By this point you should be able to see your Bare metal instances among the
|
||||
Redfish *Systems*:
|
||||
|
||||
.. code-block:: bash
|
||||
@ -355,7 +353,7 @@ Redfish *Systems*:
|
||||
"@Redfish.Copyright": "Copyright 2014-2016 Distributed Management Task Force, Inc. (DMTF). For the full DMTF copyright policy, see http://www.dmtf.org/about/policies/copyright."
|
||||
}
|
||||
|
||||
And flip its power state via the Redfish call:
|
||||
And flip an instance's power state via the Redfish call:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -382,12 +380,12 @@ Or update their boot device:
|
||||
Systems resource driver: fake
|
||||
+++++++++++++++++++++++++++++
|
||||
|
||||
The ``fake`` system driver is designed to conduct large-scale testing of
|
||||
Ironic without having a lot of bare-metal machines or being able to create a
|
||||
large number of virtual machines. When the Redfish emulator is configured with
|
||||
the ``fake`` system backend, all operations just return success. Any
|
||||
modifications are done purely in the local cache. This way, many Ironic
|
||||
operations can be tested at scale without access to a large computing pool.
|
||||
The ``fake`` system driver is designed to conduct large-scale testing of Ironic
|
||||
without having a lot of bare metal machines or being able to create a large
|
||||
number of virtual machines. When the Redfish emulator is configured with the
|
||||
``fake`` system backend, all operations just return success. Any modifications
|
||||
are done purely in the local cache. This way, many Ironic operations can be
|
||||
tested at scale without access to a large computing pool.
|
||||
|
||||
System status notifications
|
||||
~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
||||
@ -396,11 +394,11 @@ The ``fake`` driver may need to simulate components that run on the VMs to test
|
||||
an end-to-end deployment. This requires a hook interface to integrate external
|
||||
components. For instance, when testing Ironic scalability, Ironic needs to
|
||||
communicate with the Ironic Python Agent (IPA). A fake IPA can be implemented
|
||||
and synchronized with the VM status using this hook, which notifies the fake
|
||||
IPA whenever the VM status changes.
|
||||
and synchronized with the VM status using this hook, which notifies the fake IPA
|
||||
whenever the VM status changes.
|
||||
|
||||
To enable notifications, set ``external_notifier`` to ``True`` in the fake system
|
||||
object:
|
||||
To enable notifications, set ``external_notifier`` to ``True`` in the fake
|
||||
System object:
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -417,17 +415,16 @@ object:
|
||||
]
|
||||
}
|
||||
|
||||
After this, whenever the fake driver updates this system object, it will send
|
||||
an HTTP ``PUT`` request with the new system object as ``JSON`` data. The
|
||||
endpoint URL can be configured with the parameter
|
||||
``EXTERNAL_NOTIFICATION_URL``.
|
||||
After this, whenever the fake driver updates this System object, it will send an
|
||||
HTTP ``PUT`` request with the new system object as ``JSON`` data. The endpoint
|
||||
URL can be configured with the parameter ``EXTERNAL_NOTIFICATION_URL``.
|
||||
|
||||
Filtering by allowed instances
|
||||
++++++++++++++++++++++++++++++
|
||||
|
||||
It is not always desirable to manage every accessible virtual machine as a
|
||||
Redfish System, such as when an OpenStack tenant has many instances which do not
|
||||
represent virtual baremetal. In this case it is possible to specify a list of
|
||||
represent virtual bare metal. In this case it is possible to specify a list of
|
||||
UUIDs which are allowed.
|
||||
|
||||
.. code-block:: bash
|
||||
@ -447,13 +444,12 @@ UUIDs which are allowed.
|
||||
Managers resource
|
||||
-----------------
|
||||
|
||||
*Managers* are emulated based on systems: each *System* has a *Manager* with
|
||||
the same UUID. The first (alphabetically) manager will pretend to manage all
|
||||
*Managers* are emulated based on Systems: each *System* has a *Manager* with the
|
||||
same UUID. The first manager (alphabetically) will pretend to manage all
|
||||
*Chassis* and potentially other resources.
|
||||
|
||||
Managers will be revealed when querying the *Managers* resource
|
||||
directly, as well as other resources they manage or have some
|
||||
other relations.
|
||||
Managers will be revealed when querying the *Managers* resource directly, as
|
||||
well as other resources they manage or have some other relations.
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -476,9 +472,9 @@ other relations.
|
||||
Chassis resource
|
||||
----------------
|
||||
|
||||
For emulating *Chassis* resource, the user can statically configure
|
||||
one or more imaginary chassis. All existing resources (e.g. *Systems*,
|
||||
*Managers*, *Drives*) will pretend to reside in the first chassis.
|
||||
For emulating the *Chassis* resource, the user can statically configure one or
|
||||
more imaginary chassis. All existing resources (e.g. *Systems*, *Managers*,
|
||||
*Drives*) will pretend to reside in the first chassis.
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -492,9 +488,8 @@ one or more imaginary chassis. All existing resources (e.g. *Systems*,
|
||||
|
||||
By default a single chassis with be configured automatically.
|
||||
|
||||
Chassis will be revealed when querying the *Chassis* resource
|
||||
directly, as well as other resources they manage or have some
|
||||
other relations.
|
||||
Chassis will be revealed when querying the *Chassis* resource directly, as well
|
||||
as other resources they manage or have some other relations.
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -515,14 +510,14 @@ other relations.
|
||||
Indicator resource
|
||||
------------------
|
||||
|
||||
*IndicatorLED* resource is emulated as a persistent emulator database
|
||||
The *IndicatorLED* resource is emulated as a persistent emulator database
|
||||
record, observable and manageable by a Redfish client.
|
||||
|
||||
By default, *Chassis* and *Systems* resources have emulated *IndicatorLED*
|
||||
sub-resource attached and *Lit*.
|
||||
By default, the *Chassis* and *Systems* resources have emulated *IndicatorLED*
|
||||
sub-resources attached and *Lit*.
|
||||
|
||||
Non-default initial indicator state can optionally be configured
|
||||
on a per-resource basis:
|
||||
Non-default initial indicator state can optionally be configured on a
|
||||
per-resource basis:
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -557,22 +552,21 @@ Redfish client can turn *IndicatorLED* into a different state:
|
||||
Virtual media resource
|
||||
----------------------
|
||||
|
||||
Virtual Media resource is emulated as a persistent emulator database
|
||||
record, observable and manageable by a Redfish client.
|
||||
The Virtual Media resource is emulated as a persistent emulator database record,
|
||||
observable and manageable by a Redfish client.
|
||||
|
||||
By default, *VirtualMedia* resource includes two emulated removable
|
||||
devices: *Cd* and *Floppy*. Each *Manager* resource gets its own collection
|
||||
of virtual media devices as a *VirtualMedia* sub-resource.
|
||||
By default, a *VirtualMedia* resource includes two emulated removable devices:
|
||||
*Cd* and *Floppy*. Each *Manager* resource gets its own collection of virtual
|
||||
media devices as a *VirtualMedia* sub-resource.
|
||||
|
||||
If currently used *Systems* resource emulation driver supports setting
|
||||
boot image, *VirtualMedia* resource will apply inserted image onto
|
||||
all the systems being managed by this manager. Setting system boot source
|
||||
to *Cd* and boot mode to *Uefi* will cause the system to boot from
|
||||
virtual media image.
|
||||
If the currently used *Systems* resource emulation driver supports setting the
|
||||
boot image, the *VirtualMedia* resource will apply the inserted image onto all
|
||||
the systems being managed by this manager. Setting the system boot source to
|
||||
*Cd* and boot mode to *Uefi* will cause the system to boot from the virtual
|
||||
media image.
|
||||
|
||||
User can change virtual media devices and their properties through
|
||||
emulator configuration (except for the OpenStack driver which only
|
||||
supports *Cd*):
|
||||
The user can change virtual media devices and their properties through emulator
|
||||
configuration (except for the OpenStack driver which only supports *Cd*):
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -633,7 +627,7 @@ On insert the OpenStack driver will:
|
||||
* Upload the image directly to glance from the URL (long running)
|
||||
* Store the URL, image ID and volume ID in server metadata properties
|
||||
`sushy-tools-image-url`, `sushy-tools-import-image`, `sushy-tools-volume`
|
||||
* Create and attach a new volume the same size as the root disk
|
||||
* Create and attach a new volume with the same size as the root disk
|
||||
* Rebuild the server with the image, replacing the contents of the root disk
|
||||
* Delete the image
|
||||
|
||||
@ -648,19 +642,21 @@ Redfish client can eject image from virtual media device:
|
||||
|
||||
On eject the OpenStack driver will:
|
||||
|
||||
* Assume the attached volume has been rewritten with a new image (an ISO installer or IPA)
|
||||
* Detach the volume
|
||||
* Create an image from the volume (long running)
|
||||
* Store the volume image ID in server metadata property `sushy-tools-volume-image`
|
||||
* Assume the attached Volume has been rewritten with a new image (an ISO
|
||||
installer or IPA)
|
||||
* Detach the Volume
|
||||
* Create an image from the Volume (long running)
|
||||
* Store the Volume image ID in server metadata property
|
||||
`sushy-tools-volume-image`
|
||||
* Rebuild the server with the new image
|
||||
* Delete the volume
|
||||
* Delete the Volume
|
||||
* Delete the image
|
||||
|
||||
Virtual media boot
|
||||
++++++++++++++++++
|
||||
|
||||
To boot a system from a virtual media device the client first needs to figure
|
||||
out which manager is responsible for the system of interest:
|
||||
To boot a system from a virtual media device, the client first needs to figure
|
||||
out which Manager is responsible for the system of interest:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -691,7 +687,8 @@ being offered:
|
||||
},
|
||||
...
|
||||
|
||||
Knowing virtual media device name, the client can check out its present state:
|
||||
Knowing the virtual media device name, the client can check out its present
|
||||
state:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -710,9 +707,9 @@ Knowing virtual media device name, the client can check out its present state:
|
||||
"WriteProtected": false,
|
||||
...
|
||||
|
||||
Assuming `http://localhost/var/tmp/mini.iso` URL points to a bootable UEFI or
|
||||
hybrid ISO, the following Redfish REST API call will insert the image into the
|
||||
virtual CD drive:
|
||||
Assuming that the `http://localhost/var/tmp/mini.iso` URL points to a bootable
|
||||
UEFI or hybrid ISO, the following Redfish REST API call will insert the image
|
||||
into the virtual CD drive:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -741,8 +738,7 @@ Querying again, the emulator should have it in the drive:
|
||||
"WriteProtected": true,
|
||||
...
|
||||
|
||||
Next, the node needs to be configured to boot from its local CD drive
|
||||
over UEFI:
|
||||
Next, the node needs to be configured to boot from its local CD drive over UEFI:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -758,8 +754,8 @@ over UEFI:
|
||||
|
||||
.. note::
|
||||
|
||||
With the OpenStack driver the boot source is changed during insert and eject, so setting
|
||||
`BootSourceOverrideTarget` to `Cd` or `Hdd` has no effect.
|
||||
With the OpenStack driver the boot source is changed during insert and eject,
|
||||
so setting `BootSourceOverrideTarget` to `Cd` or `Hdd` has no effect.
|
||||
|
||||
By this point the system will boot off the virtual CD drive when powering it on:
|
||||
|
||||
@ -771,22 +767,20 @@ By this point the system will boot off the virtual CD drive when powering it on:
|
||||
|
||||
.. note::
|
||||
|
||||
ISO files to boot from must be UEFI-bootable, libvirtd should be running on the same
|
||||
machine with sushy-emulator.
|
||||
The ISO files to boot from must be UEFI-bootable. libvirtd should be running
|
||||
on the same machine with sushy-emulator.
|
||||
|
||||
Storage resource
|
||||
----------------
|
||||
|
||||
For emulating *Storage* resource for a System of choice, the
|
||||
user can statically configure one or more imaginary storage
|
||||
instances along with the corresponding storage controllers which
|
||||
are also imaginary.
|
||||
For emulating *Storage* resource for a System of choice, the user can statically
|
||||
configure one or more imaginary storage instances along with the corresponding
|
||||
storage controllers which are also imaginary.
|
||||
|
||||
The IDs of the imaginary drives associated to a *Storage* resource
|
||||
can be provided as a list under *Drives*.
|
||||
The IDs of the imaginary drives associated with a *Storage* resource can be
|
||||
provided as a list under *Drives*.
|
||||
|
||||
The *Storage* instances are keyed by the UUIDs of the System they
|
||||
belong to.
|
||||
The *Storage* instances are keyed by the UUIDs of the System they belong to.
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -809,8 +803,8 @@ belong to.
|
||||
]
|
||||
}
|
||||
|
||||
The Storage resources can be revealed by querying Storage resource
|
||||
for the corresponding System directly.
|
||||
The Storage resources can be revealed by querying the Storage resource for the
|
||||
corresponding System directly.
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -832,13 +826,13 @@ for the corresponding System directly.
|
||||
Drive resource
|
||||
++++++++++++++
|
||||
|
||||
For emulating the *Drive* resource, the user can statically configure
|
||||
one or more drives.
|
||||
For emulating the *Drive* resource, the user can statically configure one or
|
||||
more Drives.
|
||||
|
||||
The *Drive* instances are keyed in a composite manner using
|
||||
(System_UUID, Storage_ID) where System_UUID is the UUID of the System
|
||||
and Storage_ID is the ID of the Storage resource to which that particular
|
||||
drive belongs.
|
||||
(System_UUID, Storage_ID), where System_UUID is the UUID of the System and
|
||||
Storage_ID is the ID of the Storage resource to which that particular Drive
|
||||
belongs.
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -877,18 +871,17 @@ Storage resource it belongs to.
|
||||
Storage Volume resource
|
||||
+++++++++++++++++++++++
|
||||
|
||||
The *Volume* resource is emulated as a persistent emulator database
|
||||
record, backed by the libvirt virtualization backend of the dynamic
|
||||
Redfish emulator.
|
||||
The *Volume* resource is emulated as a persistent emulator database record,
|
||||
backed by the libvirt virtualization backend of the dynamic Redfish emulator.
|
||||
|
||||
Only the volumes specified in the config file or created via a POST request
|
||||
are allowed to be emulated upon by the emulator and appear as libvirt volumes
|
||||
in the libvirt virtualization backend. Volumes other than these can neither be
|
||||
listed nor deleted.
|
||||
Only the volumes specified in the config file or created via a POST request are
|
||||
allowed to be emulated upon by the emulator and appear as libvirt volumes in the
|
||||
libvirt virtualization backend. Volumes other than these can neither be listed
|
||||
nor deleted.
|
||||
|
||||
To allow libvirt volumes to be emulated upon, they need to be specified
|
||||
in the configuration file in the following format (keyed compositely by
|
||||
the System UUID and the Storage ID):
|
||||
To allow libvirt volumes to be emulated upon, they need to be specified in the
|
||||
configuration file in the following format (keyed compositely by the System UUID
|
||||
and the Storage ID):
|
||||
|
||||
.. code-block:: python
|
||||
|
||||
@ -913,8 +906,8 @@ the System UUID and the Storage ID):
|
||||
]
|
||||
}
|
||||
|
||||
The Volume resources can be revealed by querying Volumes resource
|
||||
for the corresponding System and the Storage.
|
||||
The Volume resources can be revealed by querying the Volumes resource for the
|
||||
corresponding System and Storage.
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -935,8 +928,8 @@ for the corresponding System and the Storage.
|
||||
"@odata.id": "/redfish/v1/Systems/da69abcc-dae0-4913-9a7b-d344043097c0/Storage/1/Volumes",
|
||||
}
|
||||
|
||||
A new volume can also be created in the libvirt backend via a POST request
|
||||
on a Volume Collection:
|
||||
A new volume can also be created in the libvirt backend via a POST request on a
|
||||
Volume Collection:
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
|
||||
@ -4,9 +4,9 @@ Using Redfish emulators
|
||||
|
||||
The sushy-tools package includes two emulators - static and dynamic.
|
||||
|
||||
Static emulator could be used to serve Redfish mocks in form of static
|
||||
JSON documents. Dynamic emulator relies upon `libvirt`, `OpenStack` or
|
||||
`Ironic` virtualization backend to mimic nodes behind a Redfish BMC.
|
||||
The static emulator can be used to serve Redfish mocks in the form of static
|
||||
JSON documents. The dynamic emulator relies upon the `libvirt`, `OpenStack` or
|
||||
`Ironic` virtualization backends to mimic nodes behind a Redfish BMC.
|
||||
|
||||
.. toctree::
|
||||
:maxdepth: 2
|
||||
|
||||
@ -2,13 +2,11 @@
|
||||
Static Redfish BMC
|
||||
==================
|
||||
|
||||
The static Redfish responder is a simple REST API server which serves
|
||||
static contents down to the Redfish client. The tool emulates the
|
||||
simple read-only BMC.
|
||||
The static Redfish responder is a simple REST API server which serves static
|
||||
contents down to the Redfish client. The tool emulates the simple read-only BMC.
|
||||
|
||||
The user is expected to supply the Redfish-compliant contents perhaps
|
||||
downloaded from the `DMTF <https://www.dmtf.org/>`_ web site. For
|
||||
example,
|
||||
The user is expected to supply the Redfish-compliant contents, perhaps
|
||||
downloaded from the `DMTF <https://www.dmtf.org/>`_ web site. For example,
|
||||
`this .zip archive <https://www.dmtf.org/sites/default/files/standards/documents/DSP2043_1.0.0.zip>`_
|
||||
includes Redfish content mocks for Redfish 1.0.0.
|
||||
|
||||
@ -19,8 +17,8 @@ includes Redfish content mocks for Redfish 1.0.0.
|
||||
unzip -d mockups DSP2043_1.0.0.zip
|
||||
sushy-static -m mockups/public-rackmount
|
||||
|
||||
Once you have the static emulator running, you can use it as it was a
|
||||
read-only bare-metal controller listening at *localhost:8000* (by default):
|
||||
Once you have the static emulator running, you can use it as if it was a
|
||||
read-only bare metal controller listening at *localhost:8000* (by default):
|
||||
|
||||
.. code-block:: bash
|
||||
|
||||
@ -39,6 +37,6 @@ read-only bare-metal controller listening at *localhost:8000* (by default):
|
||||
"@Redfish.Copyright": "Copyright 2014-2016 Distributed Management Task Force, Inc. (DMTF). For the full DMTF copyright policy, see http://www.dmtf.org/about/policies/copyright."
|
||||
}
|
||||
|
||||
You can mock different Redfish versions as well as different bare-metal
|
||||
machines by providing appropriate Redfish contents.
|
||||
You can mock different Redfish versions as well as different bare metal
|
||||
machines by providing the appropriate Redfish contents.
|
||||
|
||||
|
||||
@ -51,7 +51,7 @@ def update_service_simple_update():
|
||||
name = target.rstrip("/").rsplit('/', 1)[-1]
|
||||
uuid = flask.current_app.systems.uuid(name)
|
||||
except error.AliasAccessError as exc:
|
||||
api_utils.debug('Received a redirect in respose to GET System '
|
||||
api_utils.debug('Received a redirect in response to GET System '
|
||||
'"%s". New System ID: "%s"', name, exc)
|
||||
uuid = str(exc)
|
||||
except Exception as exc:
|
||||
|
||||
@ -211,12 +211,12 @@ class FakeDriver(AbstractSystemsDriver):
|
||||
# Check if the request was unsuccessful
|
||||
if resp.status_code >= 400:
|
||||
self._logger.error(
|
||||
'External notifcation to (%s) about system %s request'
|
||||
'External notification to (%s) about system %s request'
|
||||
'error %d: %s',
|
||||
external_notification_url, system.get('name'),
|
||||
resp.status_code, resp.text)
|
||||
return
|
||||
|
||||
# Log successful notification
|
||||
self._logger.info("External notifcation to (%s) sent about %s",
|
||||
self._logger.info("External notification to (%s) sent about %s",
|
||||
external_notification_url, system.get('name'))
|
||||
|
||||
Loading…
Reference in New Issue
Block a user