Skip to content

Commit 718ed4d

Browse files
xenolinuxxenolinux
authored
Merge pull request #96367 from xenolinux/nodepool-block-hcp
OSDOCS#14018: Configuring node pool capacity blocks in HCP
2 parents aa68961 + 0f1c480 commit 718ed4d

File tree

3 files changed

+175
-1
lines changed

3 files changed

+175
-1
lines changed

‎hosted_control_planes/hcp-manage/hcp-manage-aws.adoc‎

Lines changed: 5 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -36,4 +36,8 @@ include::modules/hcp-migrate-aws-single-to-multiarch.adoc[leveloffset=+1]
3636

3737
include::modules/hcp-migrate-aws-multiarch-nodepools.adoc[leveloffset=+1]
3838

39-
include::modules/hcp-aws-tags.adoc[leveloffset=+1]
39+
include::modules/hcp-aws-tags.adoc[leveloffset=+1]
40+
41+
include::modules/hcp-np-capacity-blocks.adoc[leveloffset=+1]
42+
43+
include::modules/hcp-np-capacity-blocks-destroy.adoc[leveloffset=+2]

‎modules/hcp-np-capacity-blocks-destroy.adoc‎

Lines changed: 28 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,28 @@
1+
// Module included in the following assemblies:
2+
// hosted_control_planes/hcp-manage/hcp-manage-aws.adoc
3+
4+
:_mod-docs-content-type: PROCEDURE
5+
[id="hcp-np-capacity-blocks-destroy_{context}"]
6+
= Destroying a hosted cluster after configuring node pool capacity blocks
7+
8+
After you configured node pool capacity blocks, you can optionally destroy a hosted cluster and uninstall the HyperShift Operator.
9+
10+
.Procedure
11+
12+
. To destroy a hosted cluster, run the following example command:
13+
+
14+
[source,terminal]
15+
----
16+
$ hcp destroy cluster aws \
17+
--name cb-np-hcp \
18+
--aws-creds $HOME/.aws/credentials \
19+
--namespace clusters \
20+
--region us-east-2
21+
----
22+
23+
. To uninstall the HyperShift Operator, run the following command:
24+
+
25+
[source,terminal]
26+
----
27+
$ hcp install render --format=yaml | oc delete -f -
28+
----

‎modules/hcp-np-capacity-blocks.adoc‎

Lines changed: 142 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,142 @@
1+
// Module included in the following assemblies:
2+
// hosted_control_planes/hcp-manage/hcp-manage-aws.adoc
3+
4+
:_mod-docs-content-type: PROCEDURE
5+
[id="hcp-np-capacity-blocks_{context}"]
6+
= Configuring node pool capacity blocks on {aws-short}
7+
8+
After creating a hosted cluster, you can configure node pool capacity blocks for graphics processing unit (GPU) reservations on {aws-first}.
9+
10+
.Procedure
11+
12+
. Create GPU reservations on {aws-short} by running the following command:
13+
+
14+
[IMPORTANT]
15+
====
16+
The zone of the GPU reservation must match your hosted cluster zone.
17+
====
18+
+
19+
[source,terminal]
20+
----
21+
$ aws ec2 describe-capacity-block-offerings \
22+
--instance-type "p4d.24xlarge"\ #<1>
23+
--instance-count "1" \ #<2>
24+
--start-date-range "$(date -u +"%Y-%m-%dT%H:%M:%SZ")" \ #<3>
25+
--end-date-range "$(date -u -d "2 day" +"%Y-%m-%dT%H:%M:%SZ")" \ #<4>
26+
--capacity-duration-hours 24 \ #<5>
27+
--output json
28+
----
29+
<1> Defines the type of your {aws-short} instance, for example, `p4d.24xlarge`.
30+
<2> Defines your instance purchase quantity, for example, `1`. Valid values are integers ranging from `1` to `64`.
31+
<3> Defines the start date range, for example, `2025-07-21T10:14:39Z`.
32+
<4> Defines the end date range, for example, `2025-07-22T10:16:36Z`.
33+
<5> Defines the duration of capacity blocks in hours, for example, `24`.
34+
35+
. Purchase the minimum fee capacity block by running the following command:
36+
+
37+
[source,terminal]
38+
----
39+
$ aws ec2 purchase-capacity-block \
40+
--capacity-block-offering-id "${MIN_FEE_ID}" \ #<1>
41+
--instance-platform "Linux/UNIX"\ #<2>
42+
--tag-specifications 'ResourceType=capacity-reservation,Tags=[{Key=usage-cluster-type,Value=hypershift-hosted}]' \ #<3>
43+
--output json > "${CR_OUTPUT_FILE}"
44+
----
45+
<1> Defines the ID of the capacity block offering.
46+
<2> Defines the platform of your instance.
47+
<3> Defines the tag for your instance.
48+
49+
. Create an environment variable to set the capacity reservation ID by running the following command:
50+
+
51+
[source,terminal]
52+
----
53+
$ CB_RESERVATION_ID=$(jq -r '.CapacityReservation.CapacityReservationId' "${CR_OUTPUT_FILE}")
54+
----
55+
+
56+
Wait for a couple of minutes for the GPU reservation to become available.
57+
58+
. Add a node pool to use the GPU reservation by running the following command:
59+
+
60+
[source,terminal]
61+
----
62+
$ hcp create nodepool aws \
63+
--cluster-name <hosted_cluster_name> \ #<1>
64+
--name <node_pool_name> \ #<2>
65+
--node-count 1 \ #<3>
66+
--instance-type p4d.24xlarge \ #<4>
67+
--arch amd64 \ #<5>
68+
--release-image <release_image> \ #<6>
69+
--render > /tmp/np.yaml
70+
----
71+
<1> Replace `<hosted_cluster_name>` with the name of your hosted cluster.
72+
<2> Replace `<node_pool_name>` with the name of your node pool.
73+
<3> Defines the node pool count, for example, `1`.
74+
<4> Defines the instance type, for example, `p4d.24xlarge`.
75+
<5> Defines an architecture type, for example, `amd64`.
76+
<6> Replace `<release_image>` with the release image you want to use.
77+
78+
. Add the `capacityReservation` setting in your `NodePool` resource by using the following example configuration:
79+
+
80+
[source,yaml]
81+
----
82+
# ...
83+
spec:
84+
arch: amd64
85+
clusterName: cb-np-hcp
86+
management:
87+
autoRepair: false
88+
upgradeType: Replace
89+
platform:
90+
aws:
91+
instanceProfile: cb-np-hcp-dqppw-worker
92+
instanceType: p4d.24xlarge
93+
rootVolume:
94+
size: 120
95+
type: gp3
96+
subnet:
97+
id: subnet-00000
98+
placement:
99+
capacityReservation:
100+
id: ${CB_RESERVATION_ID}
101+
marketType: CapacityBlocks
102+
type: AWS
103+
# ...
104+
----
105+
106+
. Apply the node pool configuration by running the following command:
107+
+
108+
[source,terminal]
109+
----
110+
$ oc apply -f /tmp/np.yaml
111+
----
112+
113+
.Verification
114+
115+
. Verify that your new node pool is created successfully by running the following command:
116+
+
117+
[source,terminal]
118+
----
119+
$ oc get np -n clusters
120+
----
121+
+
122+
.Example output
123+
[source,terminal]
124+
----
125+
NAMESPACE NAME CLUSTER DESIRED NODES CURRENT NODES AUTOSCALING AUTOREPAIR VERSION UPDATINGVERSION UPDATINGCONFIG MESSAGE
126+
clusters cb-np cb-np-hcp 1 1 False False 4.20.0-0.nightly-2025-06-05-224220 False False
127+
----
128+
129+
. Verify that your new compute nodes are created in the hosted cluster by running the following command:
130+
+
131+
[source,terminal]
132+
----
133+
$ oc get nodes
134+
----
135+
+
136+
.Example output
137+
[source, terminal]
138+
----
139+
NAME STATUS ROLES AGE VERSION
140+
ip-10-0-132-74.ec2.internal Ready worker 17m v1.32.5
141+
ip-10-0-134-183.ec2.internal Ready worker 4h5m v1.32.5
142+
----

0 commit comments

Comments
 (0)