-
Notifications
You must be signed in to change notification settings - Fork 127
Open
Description
Please select the type of request
Bug
Tell us more
Describe the request
Since release 1.27, the SOK has provided two properties on the SearchHeadCluster custom resource to control affinity and resource consumption of the Deployer:
- deployerNodeAffinity
- deployerResourceSpec
However, neither of these are exposed in the Helm chart, which means they cannot be used in a Helm-based deployment without customizing the template.
Expected behavior
Given the following values.yaml
---
# values.yaml
searchHeadCluster:
# ...
deployerNodeAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- key: app.kubernetes.io/part-of
operator: In
values:
- splunk
- key: splunk.io/instance-type
operator: In
values:
- shcd
deployerResourceSpec:
requests:
cpu: "256m"
memory: "4Gi"
limits:
cpu: "1000m"
memory: "8Gi"the properties above should be rendered into the SearchHeadCluster yaml:
---
# templated chart ouptut
apiVersion: enterprise.splunk.com/v4
kind: SearchHeadCluster
metadata:
name: sh
namespace: splunk
spec:
# ...
deployerNodeAffinity:
requiredDuringSchedulingIgnoredDuringExecution:
nodeSelectorTerms:
- matchExpressions:
- key: app.kubernetes.io/part-of
operator: In
values:
- splunk
- key: splunk.io/instance-type
operator: In
values:
- shcd
deployerResourceSpec:
limits:
cpu: 1000m
memory: 8Gi
requests:
cpu: 256m
memory: 4GiSplunk setup on K8S
- we are deploying Splunk on EKS using the
splunk-enterprisechart version 3.0.0
Reproduction/Testing steps
- rough example given in description above
K8s environment
- not relevant
Proposed changes(optional)
- update the chart template so that the fields are rendered
Metadata
Metadata
Assignees
Labels
No labels