CODEWITHSUNDEEP
databricks
munnahbaba
munnahbaba

Reputation: 251

Calling databricks notebook using Databricks Job api runs-submit endpoint

I am trying to establish an AWS lambda function which calls a databricks notebook (in the event of an s3 trigger).I understand I have to use the Jobs API of databricks in my lambda function(python) code to make a POST request using the JSON payload of the runs-submit function.

Although the documentation is not very clear, I was able to call a test script and on checking the response text I see the databricks login page html code which means it is not getting authenticated .

I did read on user tokens but I am not sure how to even incorporate them for authentication.

Any help of making this work in other ways or helping me use the user_tokens to get authenticated so that the flow reaches the execution of the notebook rather than getting stopped at authentication page would be helpful.

Thanks in advance.

Code Sample:

import requests
import json

job_payload = {
  "run_name": 'just_a_run',
  "existing_cluster_id": '****',
  "notebook_task": 
    {
      "notebook_path": 'https://databricks.cloud.company.com/****'
    }
}

resp = requests.post('https://databricks.cloud.company.com/2.0/jobs/runs/submit', json=job_payload)
print(resp.status_code)
print(resp.text)

200


<!DOCTYPE html>

<html>
<head>
    <meta charset="utf-8"/>
    <meta http-equiv="Content-Language" content="en"/>
    <title>Databricks - Sign In</title>
    <meta name="viewport" content="width=960">
    <link rel="stylesheet" href="/login/bootstrap.min.css">
    <link rel="icon" type="image/png" href="login/favicon.ico" />

    <meta http-equiv="content-type" content="text/html; charset=UTF8">
<link rel="shortcut icon" href="favicon.ico"><link href="login/login.e555bb48.css" rel="stylesheet"></head>
<body>
<div id="login-page"></div>
<script type="text/javascript" src="login/login.dabd48fd.js"></script></body>
</html>

Upvotes: 8

Views: 18247

Answers (1)

munnahbaba
munnahbaba

Reputation: 251

SOLVED:

1) You will need to create a user token for authorization and send it as 'headers' parameter while performing the REST request.

2) headers={'Authorization': 'Bearer token'} In place of token must be your actual token that you get from databricks.

3) The api link must start with /api

4) Path to the databricks notebook must be absolute path i.e. "/Users/$USER_NAME/book_name"

Final Working Code:

import requests
import json

job_payload = {
  "run_name": 'just_a_run',
  "existing_cluster_id": 'id_of_cluster',
  "notebook_task": 
    {
      "notebook_path": '/Users/username/notebook_name'
    }
}

resp = requests.post('https://databricks.cloud.company.com/api/2.0/jobs/runs/submit', json=job_payload, headers={'Authorization': 'Bearer token'})

print(resp.status_code)

print(resp.text)

Upvotes: 12

Related Questions