<@ULVA73B9P> I keep getting internal server error ...
# marvin-in-the-wildm
Marco Ruta
09/11/2024, 9:29 AM@Marvin I keep getting internal server error when i try to instantiate a k8s workpool using the rest API as follow with my prefect 2.0 server:
curl -X POST http://10.30.8.177:4200/api/work_pools/ \
-H "Content-Type: application/json" \
-d '{
"name": "test-k8s-work-pool",
"type": "kubernetes",
"base_job_template": {
"job_configuration": {
"pod_template": {
"spec": {
"containers": [
{
"name": "base",
"image": "docker.io/prefecthq/prefect:2-latest",
"env": [
{
"name": "KUBECONFIG",
"value": "kube_config.yaml"
},
{
"name": "AWS_ACCESS_KEY_ID",
"value": "minio"
},
{
"name": "AWS_SECRET_ACCESS_KEY",
"value": "minio123"
},
{
"name": "MLFLOW_S3_ENDPOINT_URL",
"value": "http://10.30.8.146:9000"
}
.... etc
{"exception_message":"Internal Server Error"}%
m
Marvin
09/11/2024, 9:30 AMGrug understand you hitting wall with 500 error when making k8s workpool on Prefect 2.0 with API call. Let's break down how to fix.
1. Let's check Prefect 2.0 docs for proper k8s workpool API setup.
2. Check GitHub issues if anyone else hit same wall.
Grug need few minutes to search.
Grug find few useful things:
1. Prefect 2.0 work pool docs mention using UI, CLI, API but not give detailed API example. Make sure all required fields are set.
2. GitHub issue #14918 show similar error when using
publish_as_work_pool()job_configurationjob_configuration2 Views