Skip to content

Commit

Permalink
Merge pull request #86660 from johnwilkins/HCIDOCS-188.1
Browse files Browse the repository at this point in the history
HCIDOCS-188: Scaling UPI cluster with BMO doesn't explain host networking
  • Loading branch information
mburke5678 authored Jan 8, 2025
2 parents 0bb92bf + 6e61cd8 commit f7c2b64
Show file tree
Hide file tree
Showing 4 changed files with 168 additions and 35 deletions.
Original file line number Diff line number Diff line change
Expand Up @@ -6,14 +6,25 @@ include::_attributes/common-attributes.adoc[]

toc::[]

After deploying a user-provisioned infrastructure cluster, you can use the Bare Metal Operator (BMO) and other metal3 components to scale bare-metal hosts in the cluster. This approach helps you to scale a user-provisioned cluster in a more automated way.
After deploying a user-provisioned infrastructure cluster, you can use the Bare Metal Operator (BMO) and other metal^3^ components to scale bare-metal hosts in the cluster. This approach helps you to scale a user-provisioned cluster in a more automated way.

include::modules/about-scaling-a-user-provisioned-installation-with-the-bare-metal-operator.adoc[leveloffset=+1]
include::modules/upi-prerequisites-for-scaling-a-upi-cluster.adoc[leveloffset=+2]
include::modules/upi-limitations-for-scaling-a-upi-cluster.adoc[leveloffset=+2]
include::modules/configuring-a-provisioning-resource-to-scale-user-provisioned-clusters.adoc[leveloffset=+1]
include::modules/upi-provisioning-new-hosts-in-a-upi-cluster.adoc[leveloffset=+1]
include::modules/upi-managing-existing-hosts-in-a-upi-cluster.adoc[leveloffset=+1]
include::modules/upi-removing-hosts-from-a-upi-cluster.adoc[leveloffset=+1]

[role="_additional-resources"]
.Additional resources

* xref:../../../installing/installing_bare_metal/ipi/ipi-install-expanding-the-cluster.adoc#preparing-the-bare-metal-node_ipi-install-expanding[Preparing the bare-metal node]
* xref:../../../installing/installing_bare_metal/ipi/ipi-install-installation-workflow.adoc#root-device-hints_ipi-install-installation-workflow[Root device hints]
* xref:../../../installing/installing_bare_metal/ipi/ipi-install-expanding-the-cluster.adoc#ipi-install-diagnosing-duplicate-mac-address_ipi-install-expanding[Diagnosing a duplicate MAC address when provisioning a new host in the cluster]
// Managing existing hosts in a user-provisioned cluster by using the BMO
include::modules/upi-managing-existing-hosts-in-a-upi-cluster.adoc[leveloffset=+1]

// Removing hosts from a user-provisioned cluster by using the BMO
include::modules/upi-removing-hosts-from-a-upi-cluster.adoc[leveloffset=+1]
Original file line number Diff line number Diff line change
Expand Up @@ -7,4 +7,4 @@
[id="about-scaling-a-user-provisioned-cluster-with-the-bare-metal-operator_{context}"]
= About scaling a user-provisioned cluster with the Bare Metal Operator

You can scale user-provisioned infrastructure clusters by using the Bare Metal Operator (BMO) and other metal3 components. User-provisioned infrastructure installations do not feature the Machine API Operator. The Machine API Operator typically manages the lifecycle of bare-metal hosts in a cluster. However, it is possible to use the BMO and other metal3 components to scale nodes in user-provisioned clusters without requiring the Machine API Operator.
You can scale user-provisioned infrastructure clusters by using the Bare Metal Operator (BMO) and other metal^3^ components. User-provisioned infrastructure installations do not feature the Machine API Operator. The Machine API Operator typically manages the lifecycle of bare-metal nodes in a cluster. However, it is possible to use the BMO and other metal^3^ components to scale nodes in user-provisioned clusters without requiring the Machine API Operator.
163 changes: 137 additions & 26 deletions modules/upi-provisioning-new-hosts-in-a-upi-cluster.adoc
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
// Module included in the following assemblies:
//
// * installing/installing_bare_metal/scaling-a-user-provisioned-cluster-with-the-bare-metal-operator.adoc
// * installing/installing_bare_metal/upi/scaling-a-user-provisioned-cluster-with-the-bare-metal-operator.adoc
:_mod-docs-content-type: PROCEDURE
[id="upi-provisioning-new-hosts-in-a-upi-cluster_{context}"]
= Provisioning new hosts in a user-provisioned cluster by using the BMO
Expand All @@ -9,7 +9,7 @@ You can use the Bare Metal Operator (BMO) to provision bare-metal hosts in a use

[NOTE]
====
To provision bare-metal hosts to the cluster by using the BMO, you must set the `spec.externallyProvisioned` specification in the `BareMetalHost` custom resource to `false`.
Provisioning bare-metal hosts to the cluster by using the BMO sets the `spec.externallyProvisioned` specification in the `BareMetalHost` custom resource to `false` by default. Do not set the `spec.externallyProvisioned` specification to `true`, because this setting results in unexpected behavior.
====

.Prerequisites
Expand All @@ -20,45 +20,153 @@ To provision bare-metal hosts to the cluster by using the BMO, you must set the
.Procedure

. Create the `Secret` CR and the `BareMetalHost` CR.
. Create a configuration file for the bare-metal node. Depending if you use either a static configuration or a DHCP server, choose one of the following example `bmh.yaml` files and configure it to your needs by replacing values in the YAML to match your environment:

.. Save the following YAML in the `bmh.yaml` file:
* To deploy with a static configuration, create the following `bmh.yaml` file:
+
[source,yaml]
----
---
apiVersion: v1
kind: Secret
metadata:
name: worker1-bmc
name: openshift-worker-<num>-network-config-secret <1>
namespace: openshift-machine-api
type: Opaque
stringData:
nmstate: | <2>
interfaces: <3>
- name: <nic1_name> <4>
type: ethernet
state: up
ipv4:
address:
- ip: <ip_address> <5>
prefix-length: 24
enabled: true
dns-resolver:
config:
server:
- <dns_ip_address> <6>
routes:
config:
- destination: 0.0.0.0/0
next-hop-address: <next_hop_ip_address> <7>
next-hop-interface: <next_hop_nic1_name> <8>
---
apiVersion: v1
kind: Secret
metadata:
name: openshift-worker-<num>-bmc-secret
namespace: openshift-machine-api
type: Opaque
data:
username: <base64_of_uid>
username: <base64_of_uid> <9>
password: <base64_of_pwd>
---
apiVersion: metal3.io/v1alpha1
kind: BareMetalHost
metadata:
name: worker1
name: openshift-worker-<num>
namespace: openshift-machine-api
spec:
online: true
bootMACAddress: <nic1_mac_address> <10>
bmc:
address: <protocol>://<bmc_url> <1>
credentialsName: "worker1-bmc"
bootMACAddress: <nic1_mac_address>
externallyProvisioned: false <2>
address: <protocol>://<bmc_url> <11>
credentialsName: openshift-worker-<num>-bmc-secret
disableCertificateVerification: false
customDeploy:
method: install_coreos
online: true
userData:
name: worker-user-data-managed
namespace: openshift-machine-api
rootDeviceHints:
deviceName: <root_device_hint> <12>
preprovisioningNetworkDataName: openshift-worker-<num>-network-config-secret
----
<1> Replace all instances of `<num>` with a unique compute node number for the bare-metal nodes in the `name`, `credentialsName`, and `preprovisioningNetworkDataName` fields.
<2> Add the NMState YAML syntax to configure the host interfaces. To configure the network interface for a newly created node, specify the name of the secret that has the network configuration. Follow the `nmstate` syntax to define the network configuration for your node. See "Preparing the bare-metal node" for details on configuring NMState syntax.
<3> Optional: If you have configured the network interface with `nmstate`, and you want to disable an interface, set `state: up` with the IP addresses set to `enabled: false`.
<4> Replace `<nic1_name>` with the name of the bare-metal node's first network interface controller (NIC).
<5> Replace `<ip_address>` with the IP address of the bare-metal node's NIC.
<6> Replace `<dns_ip_address>` with the IP address of the bare-metal node's DNS resolver.
<7> Replace `<next_hop_ip_address>` with the IP address of the bare-metal node's external gateway.
<8> Replace `<next_hop_nic1_name>` with the name of the bare-metal node's external gateway.
<9> Replace `<base64_of_uid>` and `<base64_of_pwd>` with the base64 string of the user name and password.
<10> Replace `<nic1_mac_address>` with the MAC address of the bare-metal node's first NIC. See the "BMC addressing" section for additional BMC configuration options.
<11> Replace `<protocol>` with the BMC protocol, such as IPMI, Redfish, or others. Replace `<bmc_url>` with the URL of the bare-metal node's baseboard management controller.
<12> Optional: Replace `<root_device_hint>` with a device path when specifying a root device hint. See "Root device hints" for additional details.

* When configuring the network interface with a static configuration by using `nmstate`, set `state: up` with the IP addresses set to `enabled: false`:
+
[source,yaml]
----
---
apiVersion: v1
kind: Secret
metadata:
name: openshift-worker-<num>-network-config-secret
namespace: openshift-machine-api
# ...
interfaces:
- name: <nic_name>
type: ethernet
state: up
ipv4:
enabled: false
ipv6:
enabled: false
# ...
----
<1> You can only use bare-metal host drivers that support virtual media networking booting, for example `redfish-virtualmedia` and `idrac-virtualmedia`.
<2> You must set the `spec.externallyProvisioned` specification in the `BareMetalHost` custom resource to `false`. The default value is `false`.

. Create the bare-metal host object by running the following command:
* To deploy with a DHCP configuration, create the following `bmh.yaml` file:
+
[source,yaml]
----
---
apiVersion: v1
kind: Secret
metadata:
name: openshift-worker-<num>-bmc-secret <1>
namespace: openshift-machine-api
type: Opaque
data:
username: <base64_of_uid> <2>
password: <base64_of_pwd>
---
apiVersion: metal3.io/v1alpha1
kind: BareMetalHost
metadata:
name: openshift-worker-<num>
namespace: openshift-machine-api
spec:
online: true
bootMACAddress: <nic1_mac_address> <3>
bmc:
address: <protocol>://<bmc_url> <4>
credentialsName: openshift-worker-<num>-bmc
disableCertificateVerification: false
customDeploy:
method: install_coreos
userData:
name: worker-user-data-managed
namespace: openshift-machine-api
rootDeviceHints:
deviceName: <root_device_hint> <5>
----
<1> Replace `<num>` with a unique compute node number for the bare-metal nodes in the `name` and `credentialsName` fields.
<2> Replace `<base64_of_uid>` and `<base64_of_pwd>` with the base64 string of the user name and password.
<3> Replace `<nic1_mac_address>` with the MAC address of the bare-metal node's first NIC. See the "BMC addressing" section for additional BMC configuration options.
<4> Replace `<protocol>` with the BMC protocol, such as IPMI, Redfish, or others. Replace `<bmc_url>` with the URL of the bare-metal node's baseboard management controller.
<5> Optional: Replace `<root_device_hint>` with a device path when specifying a root device hint. See "Root device hints" for additional details.
+
[IMPORTANT]
====
If the MAC address of an existing bare-metal node matches the MAC address of the bare-metal host that you are attempting to provision, then the installation will fail. If the host enrollment, inspection, cleaning, or other steps fail, the Bare Metal Operator retries the installation continuously. See "Diagnosing a duplicate MAC address when provisioning a new host in the cluster" for additional details.
====

. Create the bare-metal node by running the following command:
+
[source,terminal]
----
Expand All @@ -68,27 +176,31 @@ $ oc create -f bmh.yaml
.Example output
[source,terminal]
----
secret/worker1-bmc created
baremetalhost.metal3.io/worker1 created
secret/openshift-worker-<num>-network-config-secret created
secret/openshift-worker-<num>-bmc-secret created
baremetalhost.metal3.io/openshift-worker-<num> created
----

. Approve all certificate signing requests (CSRs).

.. Verify that the provisioning state of the host is `provisioned` by running the following command:
. Inspect the bare-metal node by running the following command:
+
[source,terminal]
----
$ oc get bmh -A
$ oc -n openshift-machine-api get bmh openshift-worker-<num>
----
+
where:
+
<num>:: Specifies the compute node number.
+
.Example output
[source,terminal]
----
NAMESPACE NAME STATE CONSUMER ONLINE ERROR AGE
openshift-machine-api controller1 externally provisioned true 5m25s
openshift-machine-api worker1 provisioned true 4m45s
NAME STATE CONSUMER ONLINE ERROR
openshift-worker-<num> provisioned true
----

. Approve all certificate signing requests (CSRs).

.. Get the list of pending CSRs by running the following command:
+
[source,terminal]
Expand Down Expand Up @@ -132,5 +244,4 @@ $ oc get nodes
NAME STATUS ROLES AGE VERSION
app1 Ready worker 47s v1.24.0+dc5a2fd
controller1 Ready master,worker 2d22h v1.24.0+dc5a2fd
----
----
21 changes: 16 additions & 5 deletions modules/upi-removing-hosts-from-a-upi-cluster.adoc
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
// Module included in the following assemblies:
//
// * installing/installing_bare_metal/scaling-a-user-provisioned-cluster-with-the-bare-metal-operator.adoc
// * installing/installing_bare_metal/upi/scaling-a-user-provisioned-cluster-with-the-bare-metal-operator.adoc
:_mod-docs-content-type: PROCEDURE
[id="upi-removing-hosts-from-a-upi-cluster_{context}"]
= Removing hosts from a user-provisioned cluster by using the BMO
Expand All @@ -15,7 +15,7 @@ You can use the Bare Metal Operator (BMO) to remove bare-metal hosts from a user
.Procedure

. Cordon and drain the host by running the following command:
. Cordon and drain the node by running the following command:
+
[source,terminal]
----
Expand Down Expand Up @@ -74,6 +74,18 @@ openshift-machine-api controller1 externally provisioned true
openshift-machine-api worker1 deprovisioning true 57m
----

. Delete the host by running the following command when the `BareMetalHost` state changes to `available`:
+
[source,terminal]
----
$ oc delete bmh -n openshift-machine-api <bmh_name>
----
+
[NOTE]
====
You can run this step without having to edit the `BareMetalHost` CR. It might take some time for the `BareMetalHost` state to change from `deprovisioning` to `available`.
====

. Delete the node by running the following command:
+
[source,terminal]
Expand All @@ -83,7 +95,7 @@ $ oc delete node <node_name>

.Verification

* Verify the node is deleted by running the following command:
* Verify that you deleted the node by running the following command:
+
[source,terminal]
----
Expand All @@ -95,5 +107,4 @@ $ oc get nodes
----
NAME STATUS ROLES AGE VERSION
controller1 Ready master,worker 2d23h v1.24.0+dc5a2fd
----
----

0 comments on commit f7c2b64

Please sign in to comment.