search
Join the Community
githubEdit

circle-playJobs

A general guide for submitting jobs via API

Once your Apps, Partitioning Engines, and Assembling Engines are ready, and you've linked them together using Job Templates, you're prepared to launch test and production jobs.

ByteNite utilizes an API to handle job requests — the Customer API— simplifying and standardizing the interaction with our server.

To begin using the API and familiarizing yourself with its endpoint, we recommend setting up a Postman collection. You can find more details at Create a Postman collection from ByteNite's OAS.

Below are the steps to configure and send a new job request, broken down and explained. You may refer to the official Jobs reference for a complete parameter list, response codes, and default code examples.

1

hashtag
Get an Access Token

Begin by obtaining a temporary access token to authenticate your requests.

POST https://api.bytenite.com/v1/auth/access_token

Main ref: Access Token

Example:

POST /auth/access_token
import requests

response = requests.post(
    "https://api.bytenite.com/v1/auth/access_token",
    json = {
        "apiKey": "ey2WmEsSMK7wdxpK5MaEHXeWCD5KEJZ79Koe68yrHL4kdnnTXT01hu2iss43BdaCCMgJ3dBh2IOVCycTt1mwkT3QR1dLxRFpK7TW7ExvcuCXio6nKsGjk9dYY8nbsFffrUVvYSQYsuQoF3NIb8sS4MDyZfOgGKZL9z8x22cwrwEck7vIokVhQ9fyWRVU2vwRiX3X4bQFuqTkkWCi5Vfy8IkGkga7ZPMPb21FxqK6cHRJ3zmI1JZZoZZxERnQcWJTpZRyCP4SNTuRm3ueVDNntFqYWYYrseNLcCIS42MpR00Z9rI9I5xxuQD6VQvHVrpOaPucg1E4Vw54xXr2LKEy9uHcM5WUQHkdfhiXo6zyVbZMrbjLpepgeS4nEja="
    }
)

token = response.json()["token"]

2

hashtag
Create a New Job

Submit a new job request using an existing job template, and give it a name.

POST https://api.bytenite.com/v1/customer/jobs

Main ref: Create

Example:

POST /customer/jobs
response = requests.post(
    "https://api.bytenite.com/v1/customer/jobs",
    headers = {
        "Authorization": token
    },
    json = {
        "name": "My job with img-gen-diffusers template",
        "templateId": "img-gen-diffusers"
    }
)

jobId = response.json()["job"]["id"]

3

hashtag
Submit a Data Source and Destination

Link a data source and destination to your job, specifying input and output options as documented in the Data Sources guide.

Connecting data sources is optional: if your app doesn't require any input data to work, or doesn't output data, you can specify a bypass data source descriptor.

PATCH https://api.bytenite.com/v1/customer/jobs/{jobId}/datasource

Main ref: Update

Example:

PATCH /customer/jobs/{jobId}/datasource
response = requests.patch(
    f"https://api.bytenite.com/v1/customer/jobs/{jobId}/datasource",
    headers = {
        "Authorization": token
    },
    json = {  
        "dataSource": {
            "dataSourceDescriptor": "url",
            "params": {
                "@type": "type.googleapis.com/bytenite.data_source.HttpDataSource",
                "url": "https://storage.googleapis.com/my-public-bucket/my-input-file.txt"
            }
        },
        "dataDestination": {
            "dataSourceDescriptor": "bucket"
        }
    }
)

4

hashtag
Submit Job Parameters

If your app, partitioner, or assembler expect parameters, provide them at this step. Parameters are organized under three keys: app, partitioner, and assembler for clarity.

If your template includes parameter schemas, the parameters you submit here will be validated immediately, and any errors will be returned.

PATCH https://api.bytenite.com/v1/customer/jobs/{jobId}/params

Main ref: Update

Example:

PATCH /customer/jobs/{jobId}/params
response = requests.patch(
    f"https://api.bytenite.com/v1/customer/jobs/{jobId}/params",
    headers = {
        "Authorization": token
    },
    json = {
        "partitioner": {
            "numImages": 20
        },
        "app": {
            "prompt": "A beautiful sunset over the jungle"
        },
        "assembler": {
            "outExtension": "jpeg"
        }
    }
)

5

hashtag
Launch the Job

Run the job, including execution configurations if needed.

Please note that you need this call to initiate the processing of your job. Without this step, your job will remain in a draft state.

POST https://api.bytenite.com/v1/customer/jobs/{jobId}/run

Main ref: Manage

Example:

PATCH /customer/jobs/{jobId}/params
response = requests.post(
    f"https://api.bytenite.com/v1/customer/jobs/{jobId}/run",
    headers = {
        "Authorization": token
    },
    json = {
        "taskTimeout": 3600,
        "jobTimeout": 86400,
        "isTestJob": True
    }
)

PreviousTemporary BucketNextCustomer API

Last updated

Was this helpful?