pydbr
Databricks client SDK for Python with command line interface for Databricks REST APIs.
{:toc}
Introduction
Pydbr (short of Python-Databricks) package provides python SDK for Databricks REST API:
- dbfs
- workspace
- jobs
- runs
The package also comes with a useful CLI which might be very helpful in automation.
Installation
$ pip install pydbr
Databricks CLI
Databricks command line client provides convenient way to interact with Databricks cluster at the command line. A very popular use of such approach in in automation tasks, like DevOps pipelines or third party workflow managers.
You can call the Databricks CLI using convenient shell command pydbr:
$ pydbr --help
or using python module:
$ python -m pydbr.cli --help
To connect to the Databricks cluster, you can supply arguments at the command line:
--bearer-token--url--cluster-id
Alternatively, you can define environment variables. Command line arguments take precedence.
export DATABRICKS_URL='https://westeurope.azuredatabricks.net/'
export DATABRICKS_BEARER_TOKEN='dapixyz89u9ufsdfd0'
export DATABRICKS_CLUSTER_ID='1234-456778-abc234'
export DATABRICKS_ORG_ID='87287878293983984'
DBFS
List DBFS items
# List items on DBFS
pydbr dbfs ls --json-indent 3 FileStore/movielens
[
{
"path": "/FileStore/movielens/ml-latest-small",
"is_dir": true,
"file_size": 0,
"is_file": false,
"human_size": "0 B"
}
]
Download file from DBFS
# Download a file and print to STDOUT
pydbr dbfs get ml-latest-small/movies.csv
Download directory from DBFS
# Download recursively entire directory and store locally
pydbr dbfs get -o ml-local ml-latest-small
Workspace
Databricks workspace contains notebooks and other items.
List workspace
####################
# List workspace
# Default path is root - '/'
$ pydbr workspace ls
# auto-add leading '/'
$ pydbr workspace ls 'Users'
# Space-indentend json output with number of spaces
$ pydbr workspace --json-indent 4 ls
# Custom indent string
$ pydbr workspace ls --json-indent='>'
Export items from Databricks workspace
#####################
# Export workspace items
# Export everything in source format using defaults: format=SOURCE, path=/
pydbr workspace export -o ./.dev/export
# Export everything in DBC format
pydbr workspace export -f DBC -o ./.dev/export.
# When path is folder, export is recursive
pydbr workspace export -o ./.dev/export-utils 'Utils'
# Export single ITEM
pydbr workspace export -o ./.dev/GetML 'Utils/Download MovieLens.py'
Runs
This command group implements the jobs/runs Databricks REST API.
Submit a notebook
Implements: https://docs.databricks.com/dev-tools/api/latest/jobs.html#runs-submit
$ pydbr runs submit "Utils/Download MovieLens"
{"run_id": 4}
You can retrieve the job information using runs get:
$ pydbr runs get 4 -i 3
If you need to pass parameters, use the --parameters or -p option and specify JSON text.
$ pydbr runs submit -p '{"run_tag":"20250103"}' "Utils/Download MovieLens"
You can refer also to parameters in JSON file:
$ pydbr runs submit -p '@params.json' "Utils/Download MovieLens"
You can use the parameters in the notebook and will also be able to see them in the run metadata:
pydbr runs get-output -i 3 8
{
"notebook_output": {
"result": "Downloaded files (tag: 20250103): README.txt, links.csv, movies.csv, ratings.csv, tags.csv",
"truncated": false
},
"error": null,
"metadata": {
"job_id": 8,
"run_id": 8,
"creator_user_name": "your.name@gmail.com",
"number_in_job": 1,
"original_attempt_run_id": null,
"state": {
"life_cycle_state": "TERMINATED",
"result_state": "SUCCESS",
"state_message": ""
},
"schedule": null,
"task": {
"notebook_task": {
"notebook_path": "/Utils/Download MovieLens",
"base_parameters": {
"run_tag": "20250103"
}
}
},
"cluster_spec": {
"existing_cluster_id": "xxxx-yyyyyy-zzzzzz"
},
"cluster_instance": {
"cluster_id": "xxxx-yyyyyy-zzzzzzzz",
"spark_context_id": "8734983498349834"
},
"overriding_parameters": null,
"start_time": 1592067357734,
"setup_duration": 0,
"execution_duration": 11000,
"cleanup_duration": 0,
"trigger": null,
"run_name": "pydbr-1592067355",
"run_page_url": "https://westeurope.azuredatabricks.net/?o=89349849834#job/8/run/1",
"run_type": "SUBMIT_RUN"
}
}
Get run metadata
Implements: Databricks REST runs/get
$ pydbr runs get -i 3 6
{
"job_id": 6,
"run_id": 6,
"creator_user_name": "your.name@gmail.com",
"number_in_job": 1,
"original_attempt_run_id": null,
"state": {
"life_cycle_state": "TERMINATED",
"result_state": "SUCCESS",
"state_message": ""
},
"schedule": null,
"task": {
"notebook_task": {
"notebook_path": "/Utils/Download MovieLens"
}
},
"cluster_spec": {
"existing_cluster_id": "xxxx-yyyyy-zzzzzz"
},
"cluster_instance": {
"cluster_id": "xxxx-yyyyy-zzzzzz",
"spark_context_id": "783487348734873873"
},
"overriding_parameters": null,
"start_time": 1592062497162,
"setup_duration": 0,
"execution_duration": 11000,
"cleanup_duration": 0,
"trigger": null,
"run_name": "pydbr-1592062494",
"run_page_url": "https://westeurope.azuredatabricks.net/?o=398348734873487#job/6/run/1",
"run_type": "SUBMIT_RUN"
}
List Runs
Implements: Databricks REST runs/list
$ pydbr runs ls
To get only the runs for a particular job:
# Get job with job-id=4
$ pydbr runs ls 4 -i 3
{
"runs": [
{
"job_id": 4,
"run_id": 4,
"creator_user_name": "your.name@gmail.com",
"number_in_job": 1,
"original_attempt_run_id": null,
"state": {
"life_cycle_state": "PENDING",
"state_message": ""
},
"schedule": null,
"task": {
"notebook_task": {
"notebook_path": "/Utils/Download MovieLens"
}
},
"cluster_spec": {
"existing_cluster_id": "xxxxx-yyyy-zzzzzzz"
},
"cluster_instance": {
"cluster_id": "xxxxx-yyyy-zzzzzzz"
},
"overriding_parameters": null,
"start_time": 1592058826123,
"setup_duration": 0,
"execution_duration": 0,
"cleanup_duration": 0,
"trigger": null,
"run_name": "pydbr-1592058823",
"run_page_url": "https://westeurope.azuredatabricks.net/?o=abcdefghasdf#job/4/run/1",
"run_type": "SUBMIT_RUN"
}
],
"has_more": false
}
Export run
Implements: Databricks REST runs/export
$ pydbr runs export --content-only 4 > .dev/run-view.html
Get run output
Implements: Databricks REST runs/get-output
$ pydbr runs get-output -i 3 6
{
"notebook_output": {
"result": "Downloaded files: README.txt, links.csv, movies.csv, ratings.csv, tags.csv",
"truncated": false
},
"error": null,
"metadata": {
"job_id": 5,
"run_id": 5,
"creator_user_name": "your.name@gmail.com",
"number_in_job": 1,
"original_attempt_run_id": null,
"state": {
"life_cycle_state": "TERMINATED",
"result_state": "SUCCESS",
"state_message": ""
},
"schedule": null,
"task": {
"notebook_task": {
"notebook_path": "/Utils/Download MovieLens"
}
},
"cluster_spec": {
"existing_cluster_id": "xxxx-yyyyy-zzzzzzz"
},
"cluster_instance": {
"cluster_id": "xxxx-yyyyy-zzzzzzz",
"spark_context_id": "8973498743973498"
},
"overriding_parameters": null,
"start_time": 1592062147101,
"setup_duration": 1000,
"execution_duration": 11000,
"cleanup_duration": 0,
"trigger": null,
"run_name": "pydbr-1592062135",
"run_page_url": "https://westeurope.azuredatabricks.net/?o=89798374987987#job/5/run/1",
"run_type": "SUBMIT_RUN"
}
}
To get only the exit output:
$ pydbr runs get-output -r 6
Downloaded files: README.txt, links.csv, movies.csv, ratings.csv, tags.csv
Python Client SDK for Databricks REST APIs
To implement your own Databricks REST API client, you can use the Python Client SDK for Databricks REST APIs.
Create Databricks connection
# Get Databricks workspace connection
dbc = pydbr.connect(
bearer_token='dapixyzabcd09rasdf',
url='https://westeurope.azuredatabricks.net')
DBFS
# Get list of items at path /FileStore
dbc.dbfs.ls('/FileStore')
# Check if file or directory exists
dbc.dbfs.exists('/path/to/heaven')
# Make a directory and it's parents
dbc.dbfs.mkdirs('/path/to/heaven')
# Delete a directory recusively
dbc.dbfs.rm('/path', recursive=True)
# Download file block starting 1024 with size 2048
dbc.dbfs.read('/data/movies.csv', 1024, 2048)
# Download entire file
dbc.dbfs.read_all('/data/movies.csv')
Databricks workspace
# List root workspace directory
dbc.workspace.ls('/')
# Check if workspace item exists
dbc.workspace.exists('/explore')
# Check if workspace item is a directory
dbc.workspace.is_directory('/')
# Export notebook in default (SOURCE) format
dbc.workspace.export('/my_notebook')
# Export notebook in HTML format
dbc.workspace.export('/my_notebook', 'HTML')
Build and publish
pip install wheel twine
python setup.py sdist bdist_wheel
python -m twine upload dist/*
Metadata
Release files for pydbr 0.0.7
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| pydbr-0.0.7.tar.gz | 17.8 kB | Without provenance | Details for pydbr-0.0.7.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| pydbr-0.0.7-py3-none-any.whl | Python 3 | none | any | Details for pydbr-0.0.7-py3-none-any.whl |
Total release size: 59.9 kB
Release files / pydbr-0.0.7.tar.gz
| Download URL | pydbr-0.0.7.tar.gz |
|---|---|
| Size | 17.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
829fb2d99c263a43629a093d77b2afeddee40e105be4f8f98920e2d0db46e56a
|
|
BLAKE2b-256 checksum How to use checksums |
09c6618f1b2cacaa50ebae807f4e862bd61e8f32b5ca0d44fb2422ce74156274
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.48.0 CPython/3.7.6rc1
|
Release files / pydbr-0.0.7-py3-none-any.whl
| Download URL | pydbr-0.0.7-py3-none-any.whl |
|---|---|
| Size | 42.1 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
56b9b848523af5bd3eb8bf7a268db39a675967449201e28a2142276c2ae66b5a
|
|
BLAKE2b-256 checksum How to use checksums |
6f0cc3c2b1e82031c75c9277903ad7f8da6462f01a6dd7c17ebcab932bbf4132
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/41.2.0 requests-toolbelt/0.9.1 tqdm/4.48.0 CPython/3.7.6rc1
|
Security for pydbr 0.0.7
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.