Metal³ introduces the concept of BareMetalHost resource, which defines a physical host and its properties. The BareMetalHost embeds two well differentiated sections, the bare metal host specification and its current status.
The BareMetalHost's spec defines the desire state of the host. It contains mainly, but not only, provisioning details.
The bmc
fields contain the connection information for the BMC
(Baseboard Management Controller) on the host.
The sub-fields are
- address -- The URL for communicating with the BMC controller, based on the provider being used. See below for more details.
- credentialsName -- A reference to a secret containing the username and password for the BMC.
- disableCertificateVerification -- A boolean to skip certificate validation when true.
BMC URLs vary based on the type of BMC and the protocol used to communicate with them.
- IPMI
ipmi://<host>:<port>
, an unadorned<host>:<port>
is also accepted and the port is optional, if using the default one (623).- The ipmi privilege level can be set from the default(
ADMINISTRATOR
) toOPERATOR
with an option URL parameterprivilegelevel
.ipmi://<host>:<port>?privilegelevel=OPERATOR
- Dell iDRAC
idrac://
(oridrac+http://
to disable TLS).idrac-virtualmedia://
to use virtual media instead of PXE for attaching the provisioning image to the host.idrac-redfish://
may be used to manage iDRAC controller with the Redfish protocol over HTTPS. The URL must also contain a path to the Redfish API system endpoint.idrac-redfish://myhost.example/redfish/v1/Systems/System.Embedded.1
- Fujitsu iRMC
irmc://<host>:<port>
, where<port>
is optional if using the default.
- HUAWEI ibmc
ibmc://<host>:<port>
(oribmc+http://<host>:<port>
to disable TLS)
- HPE iLO 4
ilo4://<host>:<port>
for iLO 4 based systems and the port is optional, if using the default one (443).
- HPE iLO 5
ilo5://<host>:<port>
for iLO 5 based systems and the port is optional, if using the default one (443).
- iLO 5 Redfish
ilo5-redfish://
(orilo5-redfish+http://
to disable TLS), the hostname or IP address, and the path to the system ID are required, for exampleilo5-redfish://myhost.example/redfish/v1/Systems/MySystemExample
- Redfish
redfish://
(orredfish+http://
to disable TLS)redfish-virtualmedia://
to use virtual media instead of PXE for attaching the provisioning image to the host.- The hostname or IP address, and the path to the system ID are
required for all variants. For example
redfish://myhost.example/redfish/v1/Systems/System.Embedded.1
orredfish://myhost.example/redfish/v1/Systems/1
A boolean indicating whether the host should be powered on (true) or off (false). Changing this value will trigger a change in power state on the physical host.
The MAC address of the NIC used for provisioning the host.
The boot mode of the host, defaults to UEFI
, can also be set
to legacy
for BIOS boot, or UEFISecureBoot
.
A reference to another resource that is using the host, it could be empty if the host is not being currently used. For example, a Machine resource when the host is being used by the machine-api.
A boolean indicating whether the host provisioning and deprovisioning are managed externally. When set:
- Power status can still be managed using the
online
field. - Hardware inventory will be monitored, but no provisioning or deprovisioning operations are performed on the host.
Holds details for the image to be deployed on a given host.
The sub-fields are
- url -- The URL of an image to deploy to the host.
- checksum -- The actual checksum or a URL to a file containing the checksum for the image at image.url.
- checksumType -- Checksum algorithms can be specified. Currently
only
md5
,sha256
,sha512
are recognized. If nothing is specifiedmd5
is assumed. - format -- This is the disk format of the image. It can be one of
raw
,qcow2
,vdi
,vmdk
,live-iso
or be left unset. Setting it to raw enables raw image streaming in Ironic agent for that image. Setting it to live-iso enables iso images to live boot without deploying to disk, in this case the checksum fields are ignored.
Even though the image sub-fields are required by Ironic,
when the host provisioning is managed externally via externallyProvisioned: true
,
and power control isn't needed, the fields can be left empty.
A reference to the Secret containing the cloudinit user data and its namespace, so it can be attached to the host before it boots for configuring different aspects of the OS (like networking, storage, ...).
A reference to the Secret containing the network configuration data (e.g. network_data.json) and its namespace, so it can be attached to the host before it boots to set network up
A human-provided string to help identify the host.
This field is deprecated. See rootDeviceHints instead.
The name of the hardware profile to use. The following are the current
supported hardwareProfile
settings and their corresponding root
devices.
hardwareProfile | Root Device |
---|---|
unknown |
/dev/sda |
libvirt |
/dev/vda |
dell |
HCTL: 0:0:0:0 |
dell-raid |
HCTL: 0:2:0:0 |
openstack |
/dev/vdb |
NOTE: These are subject to change.
This field contains the information about the RAID configuration for bare metal servers.
The sub-fields are:
- hardwareRAIDVolumes -- It contains the list of logical disks for hardware
RAID. If rootDeviceHints isn't used, the first volume is the root volume.
Furthermore, it defines the desired configuration of volume in hardware RAID.
- level -- RAID level for the logical disk. The following levels are
supported:
0
,1
,2
,5
,6
,1+0
,5+0
,6+0
. - name -- Name of the volume. Should be unique within the server. If not specified, volume name will be auto-generated.
- numberOfPhysicalDisks -- Integer, number of physical disks to use for the logical disk. Defaults to minimum number of disks required for the particular RAID level.
- rotational -- If true, select only rotational disks, if false - only solid-state and NVMe. Any disk types are used by default.
- sizeGibibytes -- Size (Integer) of the logical disk to be created in GiB. If unspecified or set to 0, the maximum capacity of disk will be used for logical disk.
- level -- RAID level for the logical disk. The following levels are
supported:
- softwareRAIDVolumes -- It contains the list of logical disks for software
RAID. If rootDeviceHints isn't used, the first volume is the root volume. If
HardwareRAIDVolumes is set this item will be invalid. The number of created
Software RAID devices must be 1 or 2. If there is only one Software RAID
device, it has to be a RAID-1. If there are two, the first one has to be a
RAID-1, while the RAID level for the second one can be 0, 1, or 1+0. As the
first RAID device will be the deployment device, enforcing a RAID-1 reduces
the risk of ending up with a non-booting node in case of a disk failure.
Furthermore, SoftwareRAIDVolume defines the desired configuration of volume
in software RAID.
- level -- RAID level for the logical disk. The following levels are
supported:
0
,1
,1+0
. - physicalDisks -- A list of device hints, the number of items should be greater than or equal to 2.
- sizeGibibytes -- Size (Integer) of the logical disk to be created in GiB. If unspecified or set to 0, the maximum capacity of disk will be used for logical disk.
- level -- RAID level for the logical disk. The following levels are
supported:
If you do not set the RAID field, we will keep the current RAID configuration.
You can set the hardwareRAIDVolume
as an empty slice to clear the hardware
RAID configuration, for example:
spec:
raid:
hardwareRAIDVolume: []
NOTE: Currently the 'raid' field is only supported by ilo5/idrac/irmc.
NOTE: Software RAID will always be deleted.
NOTE: If you got following error message:
- raid settings are defined, but the node's driver %s does not support RAID.
- node's driver %s does not support hardware RAID.
- node's driver %s does not support software RAID.
You can solve it by:
- Keep raid field is nil.
- Keep hardwareRAIDVolumes field is nil.
- Keep softwareRAIDVolumes field is nil.
If the error message you get isn't included the above, you may need to check whether the BM has a RAID controller and keep the raid field blank.
This field contains the information about the BIOS configuration for bare metal servers.
The sub-fields are:
- simultaneousMultithreadingEnabled -- Allows a single physical processor core to appear as several logical processors. This supports following options: true, false.
- sriovEnabled -- SR-IOV support enables a hypervisor to create virtual instances of a PCI-express device, potentially increasing performance. This supports following options: true, false.
- virtualizationEnabled -- Supports the virtualization of platform hardware. This supports following options: true, false.
NOTE: Currently the firmware
field is only supported by ilo4/ilo5/irmc
/idrac.
Guidance for how to choose the device to receive the image being provisioned. The storage devices are examined in the order they are discovered during inspection and the hint values are compared to the inspected values. The first discovered device that matches is used. Hints can be combined, and if multiple hints are provided then a device must match all hints in order to be selected.
The sub-fields are
- deviceName -- A string containing a Linux device name like
/dev/vda
. The hint must match the actual value exactly. - hctl -- A string containing a SCSI bus address like
0:0:0:0
. The hint must match the actual value exactly. - model -- A string containing a vendor-specific device identifier. The hint can be a substring of the actual value.
- vendor -- A string containing the name of the vendor or manufacturer of the device. The hint can be a substring of the actual value.
- serialNumber -- A string contianing the device serial number. The hint must match the actual value exactly.
- minSizeGigabytes -- An integer representing the minimum size of the device in Gigabytes.
- wwn -- A string containing the unique storage identifier. The hint must match the actual value exactly.
- wwnWithExtension -- A string containing the unique storage identifier with the vendor extension appended. The hint must match the actual value exactly.
- wwnVendorExtension -- A string containing the unique vendor storage indentifier. The hint must match the actual value exactly.
- rotational -- A boolean indicating whether the device should be
a rotating disk (
true
) or not (false
).
An interface to enable/disable automated cleaning during provisioning
and deprovisioning. When set to disabled
, automated cleaning will be
skipped, where metadata
(default value) enables it.
Moving onto the next block, the BareMetalHost's status which represents the host's current state. Including tested credentials, current hardware details, etc.
A reference to the secret and its namespace holding the last set of BMC credentials the system was able to validate as working.
A reference to the secret and its namespace holding the last set of BMC credentials that were sent to the provisioning backend.
The timestamp of the last time the status of the host was updated.
The status of the server. Value is one of the following:
- OK -- Indicates all the details for the host are known and working, meaning the host is correctly configured and manageable.
- discovered -- Implies some of the host's details are either not working correctly or missing. For example, the BMC address is known but the login credentials are not.
- error -- Indicates the system found some sort of irrecuperable error. Refer to the errorMessage field in the status section for more details.
Details of the last error reported by the provisioning backend, if any.
The details for hardware capabilities discovered on the host. These are filled in by the provisioning agent when the host is registered.
The sub-fields are
- nics -- List of network interfaces for the host.
- name -- A string identifying the network device, e.g. nic-1.
- mac -- The MAC address of the NIC.
- ip -- The IP address of the NIC, if one was assigned when the discovery agent ran.
- speedGbps -- The speed of the device in Gbps.
- vlans -- A list holding all the VLANs available for this NIC.
- vlanId -- The untagged VLAN ID.
- pxe -- Whether the NIC is able to boot using PXE.
- storage -- List of storage (disk, SSD, etc.) available to the host.
- name -- A string identifying the storage device, e.g. disk 1 (boot).
- rotational -- Either true or false, indicates whether the disk is rotational.
- sizeBytes -- Size of the storage device.
- serialNumber -- The device's serial number.
- cpu -- Details of the CPU(s) in the system.
- arch -- The architecture of the CPU.
- model -- The model string.
- clockMegahertz -- The speed in MHz of the CPU.
- flags -- List of CPU flags, e.g. 'mmx','sse','sse2','vmx', ...
- count -- Amount of these CPUs available in the system.
- firmware -- Contains BIOS information like for instance its vendor and version.
- systemVendor -- Contains information about the host's manufacturer, the productName and serialNumber.
- ramMebibytes -- The host's amount of memory in Mebibytes.
This field is deprecated. See rootDeviceHints instead.
The name of the hardware profile that matches the hardware discovered
on the host based on the details saved to the Hardware section. If
the hardware does not match any known profile, the value unknown
will be set on this field and is used by default. In practice, this
only affects which device the OS image will be written to. The
following are the current supported hardwareProfile
settings and
their corresponding root devices.
hardwareProfile | Root Device |
---|---|
unknown |
/dev/sda |
libvirt |
/dev/vda |
dell |
HCTL: 0:0:0:0 |
dell-raid |
HCTL: 0:2:0:0 |
openstack |
/dev/vdb |
NOTE: These are subject to change.
Boolean indicating whether the host is powered on.
See online on the BareMetalHost's Spec.
Settings related to deploying an image to the host.
- state -- The current state of any ongoing provisioning operation.
The following are the currently supported ones:
- <empty string> -- There is no provisioning happening, at the moment.
- registering -- The host's BMC details are being checked.
- match profile -- The discovered hardware details on the host are being compared against known profiles.
- available -- The host is available to be consumed. (This state was previously known as ready.)
- provisioning -- An image is being written to the host's disk(s).
- provisioned -- An image has been completely written to the host's disk(s).
- externally provisioned -- Metal³ does not manage the image on the host.
- deprovisioning -- The image is being wiped from the host's disk(s).
- inspecting -- The hardware details for the host are being collected by an agent.
- id -- The unique identifier for the service in the underlying provisioning tool.
- image -- The image most recently provisioned to the host.
- raid -- The list of hardware or software RAID volumes recently set.
- firmware -- The BIOS configuration for bare metal server.
- rootDeviceHints -- The root device selection instructions used for the most recent provisioning operation.
The following is a complete example from a running cluster of a BareMetalHost resource (in YAML), it includes its specification and status sections:
apiVersion: metal3.io/v1alpha1
kind: BareMetalHost
metadata:
creationTimestamp: "2019-09-20T06:33:35Z"
finalizers:
- baremetalhost.metal3.io
generation: 2
name: bmo-master-0
namespace: bmo-project
resourceVersion: "22642"
selfLink: /apis/metal3.io/v1alpha1/namespaces/bmo-project/baremetalhosts/bmo-master-0
uid: 92b2f77a-db70-11e9-9db1-525400764849
spec:
bmc:
address: ipmi://10.10.57.19
credentialsName: bmo-master-0-bmc-secret
bootMACAddress: 98:03:9b:61:80:48
consumerRef:
apiVersion: machine.openshift.io/v1beta1
kind: Machine
name: bmo-master-0
namespace: bmo-project
externallyProvisioned: true
hardwareProfile: default
image:
checksum: http://172.16.1.100/images/myOSv1/myOS.qcow2.md5sum
url: http://172.16.1.100/images/myOSv1/myOS.qcow2
online: true
raid:
hardwareRAIDVolumes:
- level: "1"
sizeGibibytes: 200
rotational: true
firmware:
virtualizationEnabled: true
userData:
name: bmo-master-user-data
namespace: bmo-project
networkData:
name: bmo-master-network-data
namespace: bmo-project
metaData:
name: bmo-master-meta-data
namespace: bmo-project
status:
errorMessage: ""
goodCredentials:
credentials:
name: bmo-master-0-bmc-secret
namespace: bmo-project
credentialsVersion: "5562"
hardware:
cpu:
arch: x86_64
clockMegahertz: 2000
count: 40
flags: []
model: Intel(R) Xeon(R) Gold 6138 CPU @ 2.00GHz
firmware:
bios:
date: 12/17/2018
vendor: Dell Inc.
version: 1.6.13
hostname: bmo-master-0.localdomain
nics:
- ip: 172.22.135.105
mac: "00:00:00:00:00:00"
model: unknown
name: eno1
pxe: true
speedGbps: 25
vlanId: 0
ramMebibytes: 0
storage: []
systemVendor:
manufacturer: Dell Inc.
productName: PowerEdge r460
serialNumber: ""
hardwareProfile: ""
lastUpdated: "2019-09-20T07:03:23Z"
operationalStatus: OK
poweredOn: true
provisioning:
ID: a4438010-3fc6-4c5c-b570-900bbe85da57
image:
checksum: ""
url: ""
state: externally provisioned
triedCredentials:
credentials:
name: bmo-master-0-bmc-secret
namespace: bmo-project
credentialsVersion: "5562"
And here is the secret bmo-master-0-bmc-secret
holding the host's
BMC credentials, base64 encoded:
$echo -n 'admin' | base64
YWRtaW4=
$echo -n 'password' | base64
cGFzc3dvcmQ=
Copy the above base64 encoded username and password pair and paste it into the yaml as mentioned below.
---
apiVersion: v1
kind: Secret
metadata:
name: bmo-master-0-bmc-secret
type: Opaque
data:
username: YWRtaW4=
password: cGFzc3dvcmQ=
NOTE: After decoding the secret content, whitespace is stripped from the beginning and end before the username and password values are used.
Several conditions must be met in order to initiate provisioning.
- The host
spec.image.url
field must contain a URL for a valid image file that is visible from within the cluster and from the host receiving the image. - The host must have
online
set totrue
so that the operator will keep the host powered on. - The host must have all of the BMC details.
To initiate deprovisioning, clear the image URL from the host spec.
Hosts created without BMC details will be left in the unmanaged
state until the details are provided. Unmanaged hosts cannot be
provisioned and their power state is undefined.
It is possible to pause the reconciliation of a BareMetalHost object by adding
an annotation baremetalhost.metal3.io/paused
. Metal³ provider sets the
value of this annotation as metal3.io/capm3
when the cluster to which the
BareMetalHost belongs, is paused and removes it when the cluster is
not paused. If you want to pause the reconciliation of BareMetalHost you can
put any value on this annotation other than metal3.io/capm3
. Please make
sure that you remove the annotation only if the value of the annotation is
not metal3.io/capm3
, but another value that you have provided. Removing the
annotation will enable the reconciliation again.
It is possible to prevent management of a BareMetalHost object by adding
an annotation baremetalhost.metal3.io/detached
. This removes the host from
the provisioner, which prevents any management of the physical host (e.g
changing power state, or deprovisioning), but still allows the BMH status
to be updated unlike the paused
anotation. While in this state the
OperationalStatus field will be detached
but the provisioning state will
be unmodified. This API only has any effect for BareMetalHost resources
that are in either Provisioned
, ExternallyProvisioned
or Ready
/Available
state.
Please note only the existence of the annotation is important to treat the BMH as detached and the value of the annotation is always ignored.