This is an automated email from the ASF dual-hosted git repository.
wilfreds pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/yunikorn-release.git
The following commit(s) were added to refs/heads/master by this push:
new 53aa084 [YUNIKORN-3212] auto update GOMEMLIMIT setting (#222)
53aa084 is described below
commit 53aa084979b8382790c3d8b7538e4a53073c326f
Author: Wilfred Spiegelenburg <[email protected]>
AuthorDate: Fri Apr 10 15:28:52 2026 +1000
[YUNIKORN-3212] auto update GOMEMLIMIT setting (#222)
Change the helm chart to calculate the GOMEMLIMIT value based on the
memory limit of the container. The goMemoryLimit value is changed to
goMemoryLimitPercentage.
Defaults:
* 80% for the scheduler container
* 60% for the admission controller and web container
The defaults are the existing fixed values converted into percentages.
Added a backwards compatible option to use the old goMemeoryLimit if
provided in the values file. The percentage version is preferred. The
fixed setting is only used if the percentage is not specified.
Updated README for helm chart
Closes: #222
Signed-off-by: Wilfred Spiegelenburg <[email protected]>
---
helm-charts/yunikorn/README.md | 125 +++++++++++----------
helm-charts/yunikorn/templates/_helpers.tpl | 30 +++++
.../templates/admission-controller-deployment.yaml | 6 +
helm-charts/yunikorn/templates/deployment.yaml | 12 ++
helm-charts/yunikorn/values.yaml | 10 +-
5 files changed, 120 insertions(+), 63 deletions(-)
diff --git a/helm-charts/yunikorn/README.md b/helm-charts/yunikorn/README.md
index 70e5ad3..9cfb3fc 100644
--- a/helm-charts/yunikorn/README.md
+++ b/helm-charts/yunikorn/README.md
@@ -64,64 +64,64 @@ helm install yunikorn yunikorn/yunikorn
## Configuration
The following table lists the configurable parameters of the YuniKorn chart
and their default values:
-| Parameter | Description
| Default |
-| ----------------------------------------------- |
----------------------------------------------------------- |
------------------------------- |
-| `imagePullSecrets` | Docker repository secrets
| `[]` |
-| `serviceAccount` | Service account name
| `yunikorn-admin` |
-| `hostNetwork` | Whether scheduler should
run in the host network | `false` |
-| `image.repository` | Scheduler image repository
| `apache/yunikorn` |
-| `image.tag` | Scheduler image tag
| `scheduler-latest` |
-| `image.pullPolicy` | Scheduler image pull
policy | `Always` |
-| `pluginImage.repository` | Scheduler plugin image
repository | `apache/yunikorn` |
-| `pluginImage.tag` | Scheduler plugin image tag
| `scheduler-plugin-latest` |
-| `pluginImage.pullPolicy` | Scheduler plugin image
pull policy | `Always` |
-| `podLabels` | Scheduler pod labels
| `{}` |
-| `podAnnotations` | Scheduler pod annotations
| `{}` |
-| `admissionController.podLabels` | Admission controller pod
labels | `{}` |
-| `admissionController.podAnnotations` | Admission controller pod
annotations | `{}` |
-| `admissionController.replicaCount` | Admission controller
replicas to be deployed | `1` |
-| `admissionController.serviceAccount` | Admission controller
service account name | `yunikorn-admission-controller` |
-| `admissionController.image.repository` | Admission controller image
repository | `apache/yunikorn` |
-| `admissionController.image.tag` | Admission controller image
tag | `admission-latest` |
-| `admissionController.image.pullPolicy` | Admission controller image
pull policy | `Always` |
-| `admissionController.hostNetwork` | Whether admission
controller should run in the host network | `true` |
-| `admissionController.resources.requests.cpu` | Admission controller CPU
resource requests | `100m` |
-| `admissionController.resources.requests.memory` | Admission controller
memory resource requests | `500Mi` |
-| `admissionController.resources.limits.cpu` | Admission controller CPU
resource limit | `500m` |
-| `admissionController.resources.limits.memory` | Admission controller
memory resource limit | `500Mi` |
-| `admissionController.goMemoryLimit` | Admission controller
memory limit (GOMEMLIMIT) | `200MiB` |
-| `admissionController.goGC` | Admission controller GC
threshold (GOGC) | `100` |
-| `admissionController.nodeSelector` | Admission controller
deployment nodeSelector(s) | `{}` |
-| `admissionController.tolerations` | Admission controller
deployment tolerations | `[]` |
-| `admissionController.affinity` | Admission controller
deployment affinity | `{}` |
-| `admissionController.service.type` | Admission controller
service type | `ClusterIP` |
-| `admissionController.priorityClassName` | Admission controller pod
priority | `""` |
-| `priorityClassName` | Scheduler pod priority
| `""` |
-| `service.type` | Scheduler service type
| `ClusterIP` |
-| `service.port` | Port of the scheduler
service | `9080` |
-| `service.portWeb` | Port of the web
application service | `9889` |
-| `resources.requests.cpu` | CPU resource requests
| `200m` |
-| `resources.requests.memory` | Memory resource requests
| `1Gi` |
-| `resources.limits.cpu` | CPU resource limit
| `4` |
-| `resources.limits.memory` | Memory resource limit
| `2Gi` |
-| `goMemoryLimit` | Memory limit (GOMEMLIMIT)
| `1536MiB` |
-| `goGC` | GC threshold (GOGC)
| `100` |
-| `web.image.repository` | Web app image repository
| `apache/yunikorn` |
-| `web.image.tag` | Web app image tag
| `web-latest` |
-| `web.image.pullPolicy` | Web app image pull policy
| `Always` |
-| `web.resources.requests.cpu` | Web app CPU resource
requests | `100m` |
-| `web.resources.requests.memory` | Web app memory resource
requests | `100Mi` |
-| `web.resources.limits.cpu` | Web app CPU resource limit
| `200m` |
-| `web.resources.limits.memory` | Web app memory resource
limit | `500Mi` |
-| `web.goMemoryLimit` | Web app memory limit
(GOMEMLIMIT) | `200MiB` |
-| `web.goGC` | Web app GC threshold
(GOGC) | `100` |
-| `embedAdmissionController` | Flag for
enabling/disabling the admission controller | `true`
|
-| `enableSchedulerPlugin` | Flag for
enabling/disabling scheduler plugin mode | `false`
|
-| `enableWebService` | Flag for
enabling/disabling web service | `true`
|
-| `nodeSelector` | Scheduler deployment
nodeSelector(s) | `{}` |
-| `tolerations` | Scheduler deployment
tolerations | `[]` |
-| `affinity` | Scheduler deployment
affinity | `{}` |
-| `yunikornDefaults` | Data for the
`yunikorn-defaults` ConfigMap | `{}`
|
+| Parameter | Description
|
Default |
+|-------------------------------------------------|-------------------------------------------------------------------------------------------------------|---------------------------------|
+| `imagePullSecrets` | Docker repository secrets
|
`[]` |
+| `serviceAccount` | Service account name
|
`yunikorn-admin` |
+| `hostNetwork` | Whether scheduler should
run in the host network |
`false` |
+| `image.repository` | Scheduler image repository
|
`apache/yunikorn` |
+| `image.tag` | Scheduler image tag
|
`scheduler-latest` |
+| `image.pullPolicy` | Scheduler image pull
policy
| `Always` |
+| `pluginImage.repository` | Scheduler plugin image
repository
| `apache/yunikorn` |
+| `pluginImage.tag` | Scheduler plugin image tag
|
`scheduler-plugin-latest` |
+| `pluginImage.pullPolicy` | Scheduler plugin image
pull policy
| `Always` |
+| `podLabels` | Scheduler pod labels
|
`{}` |
+| `podAnnotations` | Scheduler pod annotations
|
`{}` |
+| `admissionController.podLabels` | Admission controller pod
labels |
`{}` |
+| `admissionController.podAnnotations` | Admission controller pod
annotations |
`{}` |
+| `admissionController.replicaCount` | Admission controller
replicas to be deployed
| `1` |
+| `admissionController.serviceAccount` | Admission controller
service account name
| `yunikorn-admission-controller` |
+| `admissionController.image.repository` | Admission controller image
repository |
`apache/yunikorn` |
+| `admissionController.image.tag` | Admission controller image
tag |
`admission-latest` |
+| `admissionController.image.pullPolicy` | Admission controller image
pull policy |
`Always` |
+| `admissionController.hostNetwork` | Whether admission
controller should run in the host network
| `true` |
+| `admissionController.resources.requests.cpu` | Admission controller CPU
resource requests |
`100m` |
+| `admissionController.resources.requests.memory` | Admission controller
memory resource requests
| `500Mi` |
+| `admissionController.resources.limits.cpu` | Admission controller CPU
resource limit |
`500m` |
+| `admissionController.resources.limits.memory` | Admission controller
memory resource limit
| `500Mi` |
+| `admissionController.goMemoryLimitPercentage` | Percentage to calculate
the GOMEMLIMIT value with based on the container's `resources.limits.memory` |
`60` |
+| `admissionController.goGC` | Admission controller GC
threshold (GOGC) |
`100` |
+| `admissionController.nodeSelector` | Admission controller
deployment nodeSelector(s)
| `{}` |
+| `admissionController.tolerations` | Admission controller
deployment tolerations
| `[]` |
+| `admissionController.affinity` | Admission controller
deployment affinity
| `{}` |
+| `admissionController.service.type` | Admission controller
service type
| `ClusterIP` |
+| `admissionController.priorityClassName` | Admission controller pod
priority |
`""` |
+| `priorityClassName` | Scheduler pod priority
|
`""` |
+| `service.type` | Scheduler service type
|
`ClusterIP` |
+| `service.port` | Port of the scheduler
service
| `9080` |
+| `service.portWeb` | Port of the web
application service
| `9889` |
+| `resources.requests.cpu` | CPU resource requests
|
`200m` |
+| `resources.requests.memory` | Memory resource requests
|
`1Gi` |
+| `resources.limits.cpu` | CPU resource limit
|
`4` |
+| `resources.limits.memory` | Memory resource limit
|
`2Gi` |
+| `goMemoryLimitPercentage` | Percentage to calculate
the GOMEMLIMIT value with based on the container's `resources.limits.memory` |
`80` |
+| `goGC` | GC threshold (GOGC)
|
`100` |
+| `web.image.repository` | Web app image repository
|
`apache/yunikorn` |
+| `web.image.tag` | Web app image tag
|
`web-latest` |
+| `web.image.pullPolicy` | Web app image pull policy
|
`Always` |
+| `web.resources.requests.cpu` | Web app CPU resource
requests
| `100m` |
+| `web.resources.requests.memory` | Web app memory resource
requests |
`100Mi` |
+| `web.resources.limits.cpu` | Web app CPU resource limit
|
`200m` |
+| `web.resources.limits.memory` | Web app memory resource
limit |
`500Mi` |
+| `web.goMemoryLimitPercentage` | Percentage to calculate
the GOMEMLIMIT value with based on the container's `resources.limits.memory` |
`60` |
+| `web.goGC` | Web app GC threshold
(GOGC)
| `100` |
+| `embedAdmissionController` | Flag for
enabling/disabling the admission controller
| `true` |
+| `enableSchedulerPlugin` | Flag for
enabling/disabling scheduler plugin mode
| `false` |
+| `enableWebService` | Flag for
enabling/disabling web service
| `true` |
+| `nodeSelector` | Scheduler deployment
nodeSelector(s)
| `{}` |
+| `tolerations` | Scheduler deployment
tolerations
| `[]` |
+| `affinity` | Scheduler deployment
affinity
| `{}` |
+| `yunikornDefaults` | Data for the
`yunikorn-defaults` ConfigMap
| `{}` |
These parameters can be passed in via helm's `--set` and `--values (-f)`
options, such as `--set resources.requests.cpu=2000m` or `-f myvalues.yaml`.
@@ -142,3 +142,12 @@ The following configuration parameters are deprecated and
will be ignored in a f
| `configuration` | YAML-formatted queue configuration
| `queues.yaml` ConfigMap entry
|
Currently, if both the deprecated parameter and the replacement ConfigMap
entry are specified, the ConfigMap entry will take precedence.
+
+The following settings have been deprecated. The new percentage based settings
have precedence over the deprecated settings.
+The deprecated settings will only be used as a fallback if the new percentage
based settings are not present.
+
+| Parameter | Description
| Replacement |
+|-------------------------------------|------------------------------------------------|-----------------------------------------------|
+| `goMemoryLimit` | Memory limit (GOMEMLIMIT)
| `goMemoryLimitPercentage` |
+| `web.goMemoryLimit` | Web app memory limit (GOMEMLIMIT)
| `web.goMemoryLimitPercentage` |
+| `admissionController.goMemoryLimit` | Admission controller memory limit
(GOMEMLIMIT) | `admissionController.goMemoryLimitPercentage` |
diff --git a/helm-charts/yunikorn/templates/_helpers.tpl
b/helm-charts/yunikorn/templates/_helpers.tpl
index 51142cb..bcb5fe2 100644
--- a/helm-charts/yunikorn/templates/_helpers.tpl
+++ b/helm-charts/yunikorn/templates/_helpers.tpl
@@ -46,3 +46,33 @@ Create chart name and version as used by the chart label.
{{- define "yunikorn.chart" -}}
{{- printf "%s-%s" .Chart.Name .Chart.Version | replace "+" "_" | trunc 63 |
trimSuffix "-" -}}
{{- end -}}
+
+{{/*
+Convert the Kubernetes memory string (e.g. 1500Mi) into bytes: suffix dicts
merged, then value × multiplier (Ki..Ei use 0x1pN).
+*/}}
+{{- define "yunikorn.resource-quantity" -}}
+ {{- $value := . -}}
+ {{- $unit := 1.0 -}}
+ {{- $mem := . | toString | trim -}}
+ {{- if typeIs "string" . -}}
+ {{- $base2 := dict "Ki" 0x1p10 "Mi" 0x1p20 "Gi" 0x1p30 "Ti" 0x1p40
"Pi" 0x1p50 "Ei" 0x1p60 -}}
+ {{- $base10 := dict "m" 1e-3 "k" 1e3 "M" 1e6 "G" 1e9 "T" 1e12 "P" 1e15
"E" 1e18 -}}
+ {{- range $k, $v := merge $base2 $base10 -}}
+ {{- if hasSuffix $k $mem -}}
+ {{- $value = trimSuffix $k $mem -}}
+ {{- $unit = $v -}}
+ {{- end -}}
+ {{- end -}}
+ {{- end -}}
+ {{- mulf (float64 $value) $unit -}}
+{{- end -}}
+
+{{/*
+Soft memory limit: limits.memory (bytes) × (goMemoryLimitPercentage/100),
rounded up (ceil).
+*/}}
+{{- define "yunikorn.goMemLimit" -}}
+ {{- $multi := divf (float64 .goMemoryLimitPercentage) 100.0 -}}
+ {{- with .resources.limits.memory -}}
+ {{- include "yunikorn.resource-quantity" . | float64 | mulf $multi |
ceil | int -}}
+ {{- end }}
+{{- end -}}
diff --git
a/helm-charts/yunikorn/templates/admission-controller-deployment.yaml
b/helm-charts/yunikorn/templates/admission-controller-deployment.yaml
index 2dd3be3..c37c431 100644
--- a/helm-charts/yunikorn/templates/admission-controller-deployment.yaml
+++ b/helm-charts/yunikorn/templates/admission-controller-deployment.yaml
@@ -106,7 +106,13 @@ spec:
fieldRef:
fieldPath: metadata.namespace
- name: GOMEMLIMIT
+ {{- if .Values.admissionController.goMemoryLimitPercentage }}
+ {{- with include "yunikorn.goMemLimit" .Values.admissionController
}}
+ value: {{ quote . }}
+ {{- end }}
+ {{- else }}
value: "{{ .Values.admissionController.goMemoryLimit }}"
+ {{- end }}
- name: GOGC
value: "{{ .Values.admissionController.goGC }}"
ports:
diff --git a/helm-charts/yunikorn/templates/deployment.yaml
b/helm-charts/yunikorn/templates/deployment.yaml
index 0d02205..4397f35 100644
--- a/helm-charts/yunikorn/templates/deployment.yaml
+++ b/helm-charts/yunikorn/templates/deployment.yaml
@@ -98,7 +98,13 @@ spec:
fieldRef:
fieldPath: metadata.namespace
- name: GOMEMLIMIT
+ {{- if .Values.goMemoryLimitPercentage }}
+ {{- with include "yunikorn.goMemLimit" .Values }}
+ value: {{ quote . }}
+ {{- end }}
+ {{- else }}
value: "{{ .Values.goMemoryLimit }}"
+ {{- end }}
- name: GOGC
value: "{{ .Values.goGC }}"
resources:
@@ -117,7 +123,13 @@ spec:
protocol: TCP
env:
- name: GOMEMLIMIT
+ {{- if .Values.web.goMemoryLimitPercentage }}
+ {{- with include "yunikorn.goMemLimit" .Values.web }}
+ value: {{ quote . }}
+ {{- end }}
+ {{- else }}
value: "{{ .Values.web.goMemoryLimit }}"
+ {{- end }}
- name: GOGC
value: "{{ .Values.web.goGC }}"
resources:
diff --git a/helm-charts/yunikorn/values.yaml b/helm-charts/yunikorn/values.yaml
index 3d53984..679d016 100644
--- a/helm-charts/yunikorn/values.yaml
+++ b/helm-charts/yunikorn/values.yaml
@@ -57,7 +57,7 @@ admissionController:
limits:
cpu: 500m
memory: 500Mi
- goMemoryLimit: 200MiB
+ goMemoryLimitPercentage: 60
goGC: 100
nodeSelector: {}
tolerations: []
@@ -83,7 +83,7 @@ web:
limits:
cpu: 200m
memory: 500Mi
- goMemoryLimit: 200MiB
+ goMemoryLimitPercentage: 60
goGC: 100
service:
@@ -109,9 +109,9 @@ resources:
cpu: 4
memory: 2Gi
-# Go memory limit (GOMEMLIMIT) for the scheduler. This should be updated in
tandem with
-# resources.limits.memory to avoid OOMKilled scenarios.
-goMemoryLimit: 1536MiB
+# Percentage of the real memory limit of the container to which the GOMEMLIMIT
variable is set
+# GOMEMLIMIT variable sets a soft memory limit for the runtime to use in GC
runs.
+goMemoryLimitPercentage: 80
# Go garbage-collection threshold (GOGC) as a percentage of freshly allocated
data to
# live data remaining after the previous GC.
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
