Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
75 changes: 75 additions & 0 deletions modules/nodes-nodes-psi-enable.adoc
Original file line number Diff line number Diff line change
@@ -0,0 +1,75 @@
// Module included in the following assemblies:
//
// * nodes/nodes/nodes-nodes-managing.adoc

:_mod-docs-content-type: PROCEDURE
[id="nodes-nodes-psi-enable_{context}"]
= Enabling Pressure Stall Information (PSI) monitoring

[role="_abstract"]
You can enable Linux Pressure Stall Information monitoring by using a `MachineConfig` object. Enabling PSI monitoring makes PSI metrics (pressure stall information) for CPU, memory, and IO available for your cluster.

[NOTE]
====
The performance impact of enabling PSI is negligible with the benefit of gaining critical resource pressure visibility for better scheduling and eviction decisions. This was tested on worker nodes (m5.2xl) on the AWS platform. For latency sensitive environments, we suggest testing it before enabling it on production. Enable it on 1 worker first, monitor for the performance impact, and gradually roll out to more workers.
Copy link

@qiliRedHat qiliRedHat Jan 16, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The performance impact to of enabling PSI is negligible with the benefit of gaining critical resource pressure visibility for better scheduling and eviction decisions. This was tested on worker nodes (m5.2xl) on the AWS platform. For latency sensitive environments, we suggest testing it before enabling it on production. Enable it on 1 worker first, monitor for the performance impact, and gradually roll out to more workers. Enabling PSI introduces new metrics per container that can increase the RSS memory usage of prometheus-k8s pods. On a cluster with a big number of containers, monitor the memory usage of prometheus-k8s pods and adjust the resource if needed.

====

.Prerequisites

. Obtain the label associated with the static `MachineConfigPool` CR for the type of node you want to configure.
Perform one of the following steps:

.. View the machine config pool:
+
[source,terminal]
----
$ oc describe machineconfigpool <name>
----
+
For example:
+
[source,terminal]
----
$ oc describe machineconfigpool worker
----
+
.Example output
[source,yaml]
----
apiVersion: machineconfiguration.openshift.io/v1
kind: MachineConfigPool
metadata:
creationTimestamp: 2019-02-08T14:52:39Z
generation: 1
labels:
custom-kubelet: set-kubelet-config <1>
----
<1> If a label has been added it appears under `labels`.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🤖 [error] AsciiDocDITA.CalloutList: Callouts are not supported in DITA.


.. If the label is not present, add a key/value pair:
+
[source,terminal]
----
$ oc label machineconfigpool worker custom-kubelet=set-kubelet-config
----

.Procedure

. Create a YAML file similar to the following that contains the machine configuration:
+
[source,yaml]
----
apiVersion: machineconfiguration.openshift.io/v1
kind: MachineConfig
metadata:
labels:
machineconfiguration.openshift.io/role: <label>
name: 99-openshift-machineconfig-<label>-kargs
spec:
kernelArguments:
- psi=1
----
where:
+
`spec.kernelArguments.psi=1` Specifies that Pressure Stall Information (PSI) monitoring is enabled.

2 changes: 2 additions & 0 deletions nodes/nodes/nodes-nodes-managing.adoc
Original file line number Diff line number Diff line change
Expand Up @@ -34,6 +34,8 @@ include::modules/nodes-nodes-parallel-container-pulls-configure.adoc[leveloffset

include::modules/nodes-control-plane-osp-migrating.adoc[leveloffset=+1]

include::modules/nodes-nodes-psi-enable.adoc[leveloffset=+1]

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

Expand Down