Skip to content

Commit bf3ab97

Browse files
xenolinuxxenolinux
committed
OCPBUGS#50642: NTP server configuration for HCP
1 parent ee2e3c2 commit bf3ab97

File tree

2 files changed

+183
-0
lines changed

2 files changed

+183
-0
lines changed

‎hosted_control_planes/hcp-machine-config.adoc

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -27,3 +27,11 @@ include::modules/hcp-kubeconf-nodepool.adoc[leveloffset=+1]
2727
include::modules/node-tuning-hosted-cluster.adoc[leveloffset=+1]
2828

2929
include::modules/sriov-operator-hosted-control-planes.adoc[leveloffset=+1]
30+
31+
include::modules/hcp-configure-ntp.adoc[leveloffset=+1]
32+
33+
[role="_additional-resources"]
34+
.Additional resources
35+
36+
* xref:../installing/install_config/installing-customizing.adoc#installation-special-config-butane_installing-customizing[Creating machine configs with Butane]
37+
* link:https://docs.redhat.com/en/documentation/red_hat_advanced_cluster_management_for_kubernetes/2.13/html-single/clusters/index#create-host-inventory-cli-steps[Creating a host inventory]

‎modules/hcp-configure-ntp.adoc

Lines changed: 175 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,175 @@
1+
// Module included in the following assemblies:
2+
// * hosted_control_planes/hcp-machine-config.adoc
3+
4+
:_mod-docs-content-type: PROCEDURE
5+
[id="hcp-configure-ntp_{context}"]
6+
= Configuring the NTP server for hosted clusters
7+
8+
You can configure the Network Time Protocol (NTP) server for your hosted clusters by using Butane.
9+
10+
.Procedure
11+
12+
. Create a Butane config file, `99-worker-chrony.bu`, that includes the contents of the `chrony.conf` file. For more information about Butane, see "Creating machine configs with Butane".
13+
+
14+
.Example `99-worker-chrony.bu` configuration
15+
[source,yaml,subs="attributes+"]
16+
----
17+
# ...
18+
variant: openshift
19+
version: {product-version}.0
20+
metadata:
21+
name: 99-worker-chrony
22+
labels:
23+
machineconfiguration.openshift.io/role: worker
24+
storage:
25+
files:
26+
- path: /etc/chrony.conf
27+
mode: 0644 #<1>
28+
overwrite: true
29+
contents:
30+
inline: |
31+
pool 0.rhel.pool.ntp.org iburst #<2>
32+
driftfile /var/lib/chrony/drift
33+
makestep 1.0 3
34+
rtcsync
35+
logdir /var/log/chrony
36+
# ...
37+
----
38+
<1> Specify an octal value mode for the `mode` field in the machine config file. After creating the file and applying the changes, the `mode` field is converted to a decimal value.
39+
<2> Specify any valid, reachable time source, such as the one provided by your Dynamic Host Configuration Protocol (DHCP) server.
40+
+
41+
[NOTE]
42+
====
43+
For machine-to-machine communication, the NTP on the User Datagram Protocol (UDP) port is `123`. If you configured an external NTP time server, you must open UDP port `123`.
44+
====
45+
46+
. Use Butane to generate a `MachineConfig` object file, `99-worker-chrony.yaml`, that contains a configuration that Butane sends to the nodes. Run the following command:
47+
+
48+
[source,terminal]
49+
----
50+
$ butane 99-worker-chrony.bu -o 99-worker-chrony.yaml
51+
----
52+
+
53+
.Example `99-worker-chrony.yaml` configuration
54+
[source,yaml]
55+
----
56+
# Generated by Butane; do not edit
57+
apiVersion: machineconfiguration.openshift.io/v1
58+
kind: MachineConfig
59+
metadata:
60+
labels:
61+
machineconfiguration.openshift.io/role: worker
62+
name: <machineconfig_name>
63+
spec:
64+
config:
65+
ignition:
66+
version: 3.2.0
67+
storage:
68+
files:
69+
- contents:
70+
source: data:...
71+
mode: 420
72+
overwrite: true
73+
path: /example/path
74+
----
75+
76+
. Add the contents of the `99-worker-chrony.yaml` file inside of a config map in the management cluster:
77+
+
78+
.Example config map
79+
[source,yaml]
80+
----
81+
apiVersion: v1
82+
kind: ConfigMap
83+
metadata:
84+
name: <configmap_name>
85+
namespace: <namespace> #<1>
86+
data:
87+
config: |
88+
apiVersion: machineconfiguration.openshift.io/v1
89+
kind: MachineConfig
90+
metadata:
91+
labels:
92+
machineconfiguration.openshift.io/role: worker
93+
name: <machineconfig_name>
94+
spec:
95+
config:
96+
ignition:
97+
version: 3.2.0
98+
storage:
99+
files:
100+
- contents:
101+
source: data:...
102+
mode: 420
103+
overwrite: true
104+
path: /example/path
105+
# ...
106+
----
107+
<1> Replace `<namespace>` with the name of your namespace where you created the node pool, such as `clusters`.
108+
109+
. Apply the config map to your node pool by running the following command:
110+
+
111+
[source,terminal]
112+
----
113+
$ oc edit nodepool <nodepool_name> --namespace <hosted_cluster_namespace>
114+
----
115+
+
116+
.Example `NodePool` configuration
117+
[source,yaml]
118+
----
119+
apiVersion: hypershift.openshift.io/v1alpha1
120+
kind: NodePool
121+
metadata:
122+
# ...
123+
name: nodepool-1
124+
namespace: clusters
125+
# ...
126+
spec:
127+
config:
128+
- name: <configmap_name> #<1>
129+
# ...
130+
----
131+
<1> Replace `<configmap_name>` with the name of your config map.
132+
133+
. Add the list of your NTP servers in the `infra-env.yaml` file, which defines the `InfraEnv` custom resource (CR):
134+
+
135+
.Example `infra-env.yaml` file
136+
[source,yaml]
137+
----
138+
apiVersion: agent-install.openshift.io/v1beta1
139+
kind: InfraEnv
140+
# ...
141+
spec:
142+
additionalNTPSources:
143+
- <ntp_server> #<1>
144+
- <ntp_server1>
145+
- <ntp_server2>
146+
# ...
147+
----
148+
<1> Replace `<ntp_server>` with the name of your NTP server. For more details about creating a host inventory and the `InfraEnv` CR, see "Creating a host inventory".
149+
150+
. Apply the `InfraEnv` CR by running the following command:
151+
+
152+
[source,terminal]
153+
----
154+
$ oc apply -f infra-env.yaml
155+
----
156+
157+
.Verification
158+
159+
* Check the following fields to know the status of your host inventory:
160+
+
161+
** `conditions`: The standard Kubernetes conditions indicating if the image was created successfully.
162+
** `isoDownloadURL`: The URL to download the Discovery Image.
163+
** `createdTime`: The time at which the image was last created. If you modify the `InfraEnv` CR, ensure that you have updated the timestamp before downloading a new image.
164+
+
165+
Verify that your host inventory is created by running the following command:
166+
+
167+
[source,terminal]
168+
----
169+
$ oc describe infraenv <infraenv_resource_name> -n <infraenv_namespace>
170+
----
171+
+
172+
[NOTE]
173+
====
174+
If you modify the `InfraEnv` CR, confirm that the `InfraEnv` CR has created a new Discovery Image by looking at the `createdTime` field. If you already booted hosts, boot them again with the latest Discovery Image.
175+
====

0 commit comments

Comments
 (0)