Netflix Conductor not retrying failed HTTP call? - netflix-conductor

Question

I have a conductor workflow which makes an HTTP call to a service. When the service is completely down (not started), conductor just set the task and the workflow states to FAILED. I have added retry parameters to the HTTP task and am expecting the tool to retry the call 5 times before failing. I am doing anything wrong?
I am running conductor locally using docker.
Here is my workflow:
{
"createTime": 1674599429945,
"updateTime": 1674599615808,
"accessPolicy": {},
"name": "make_api_call",
"description": "A workflow that makes an api call",
"version": 2,
"tasks": [
{
"name": "make_call",
"taskReferenceName": "make_call_ref",
"inputParameters": {
"http_request": {
"uri": "${workflow.input.API_ENDPOINT}",
"method": "POST",
"headers": {
"Authorization": "Bearer ${workflow.input.token}"
},
"body": {
"email": "${workflow.input.email}"
},
"connectionTimeOut": 5000,
"readTimeOut": 5000
}
},
"type": "HTTP",
"startDelay": 0,
"optional": false,
"asyncComplete": false,
"retryCount": 5,
"timeoutSeconds": 30,
"timeoutPolicy": "RETRY"
}
],
"inputParameters": [],
"outputParameters": {
"data": {
"response": "${make_call_ref.output.response.body}"
}
},
"schemaVersion": 2,
"restartable": true,
"workflowStatusListenerEnabled": true,
"ownerEmail": "example#email.com",
"timeoutPolicy": "TIME_OUT_WF",
"timeoutSeconds": 5000,
"variables": {},
"inputTemplate": {}
}
and the input json:
{
"API_ENDPOINT": "http://host.docker.internal:4555/v1/users",
"token": "xxx",
"email": "johndoe#my-app.com"
}