[SPARK-54553][VOLCANO] Supports receiving podgroup JSON format configurations when using Volcano.

[ruanwenjun]

What changes were proposed in this pull request?

Add a new configuration option spark.kubernetes.scheduler.volcano.podGroupTemplateJson for cases where Volcano is used as the Kubernetes scheduler.
This option allows users to provide the PodGroup configuration directly as a JSON string.
If podGroupTemplateFile is not specified, Spark will fall back to using podGroupTemplateJson.

Why are the changes needed?

• Avoid creating temporary podgroup files to facilitate usage and integration with external platforms such as kyuubi.

Does this PR introduce any user-facing change?

Yes.

Use below conf can work

--conf spark.kubernetes.driver.pod.featureSteps=org.apache.spark.deploy.k8s.features.VolcanoFeatureStep \
--conf spark.kubernetes.executor.pod.featureSteps=org.apache.spark.deploy.k8s.features.VolcanoFeatureStep \
--conf spark.kubernetes.scheduler.name=volcano \
--conf spark.kubernetes.scheduler.volcano.podGroupTemplateJson={"spec": {"minMember": 1,"minResources": {"cpu": "2","memory": "3Gi"},"queue": "bigdata"}} \

The podgroup has been created correctly

Name:         spark-40cd4561661d4298a3f08cc7477cdc3c-podgroup
Namespace:    spark
API Version:  scheduling.volcano.sh/v1beta1
Kind:         PodGroup
Metadata:
  Creation Timestamp:  2025-12-05T09:18:35Z
  Generation:          7
  Owner References:
    API Version:     v1
    Controller:      true
    Kind:            Pod
    Name:            kyuubi-cql-test-ds-35faf46f-c4f9-44d3-8288-d1875e8a074d-driver
    UID:             8fd2d875-9c17-4bf2-8787-fdd4f9e2e79b
  Resource Version:  2230406043
  UID:               adaeb0a8-c6f6-4c29-92e8-480d8e01d89a
Spec:
  Min Member:  1
  Min Resources:
    Cpu:     2
    Memory:  3Gi
  Queue:     bigdata

How was this patch tested?

• Test by UT

Was this patch authored or co-authored using generative AI tooling?

NO