Skip to main content

S3 Data Lake

Module s3


Important Capabilities

Data ProfilingOptionally enabled via configuration
Extract TagsCan extract S3 object/bucket tags if enabled

This plugin extracts:

  • Row and column counts for each table
  • For each column, if profiling is enabled:
    • null counts and proportions
    • distinct counts and proportions
    • minimum, maximum, mean, median, standard deviation, some quantile values
    • histograms or frequencies of unique values

This connector supports both local files as well as those stored on AWS S3 (which must be identified using the prefix s3://). Supported file types are as follows:

  • CSV
  • TSV
  • JSON
  • Parquet
  • Apache Avro

Schemas for Parquet and Avro files are extracted as provided.

Schemas for schemaless formats (CSV, TSV, JSON) are inferred. For CSV and TSV files, we consider the first 100 rows by default, which can be controlled via the max_rows recipe parameter (see below) JSON file schemas are inferred on the basis of the entire file (given the difficulty in extracting only the first few objects of the file), which may impact performance. We are working on using iterator-based JSON parsers to avoid reading in the entire JSON object.

Note that because the profiling is run with PySpark, we require Spark 3.0.3 with Hadoop 3.2 to be installed (see compatibility for more details). If profiling, make sure that permissions for s3a:// access are set because Spark and Hadoop use the s3a:// protocol to interface with AWS (schema inference outside of profiling requires s3:// access). Enabling profiling will slow down ingestion runs.

CLI based Ingestion

Install the Plugin

pip install 'acryl-datahub[s3]'

Starter Recipe

Check out the following recipe to get started with ingestion! See below for full configuration options.

For general pointers on writing and running a recipe, see our main recipe guide.

type: s3
include: "s3://covid19-lake/covid_knowledge_graph/csv/nodes/*.*"

aws_access_key_id: *****
aws_secret_access_key: *****
aws_region: us-east-2
env: "PROD"
enabled: false

# sink configs

Config Details

Note that a . is used to denote nested fields in the YAML recipe.

View All Configuration Options
envstringThe environment that all assets produced by this connector belong toPROD
platformstringThe platform that this source connects to
platform_instancestringThe instance of the platform that all assets produced by this recipe belong toNone
path_specsArray of objectList of PathSpec. See below the details about PathSpecNone
use_s3_bucket_tagsbooleanWhether or not to create tags in datahub from the s3 bucketNone
use_s3_object_tagsboolean# Whether or not to create tags in datahub from the s3 objectNone
spark_driver_memorystringMax amount of memory to grant Spark.4g
max_rowsintegerMaximum number of rows to use when inferring schemas for TSV and CSV files.100
path_specPathSpec (see below for fields)Path spec will be deprecated in favour of path_specs option.
path_spec.include❓ (required if path_spec is set)stringPath to table (s3 or local file system). Name variable {table} is used to mark the folder with dataset. In absence of {table}, file level dataset will be created. Check below examples for more details.None
path_spec.excludeArray of stringlist of paths in glob pattern which will be excluded while scanning for the datasetsNone
path_spec.file_typesArray of stringFiles with extenstions specified here (subset of default value) only will be scanned to create dataset. Other files will be omitted.['csv', 'tsv', 'json', 'parquet', 'avro']
path_spec.default_extensionstringFor files without extension it will assume the specified file type. If it is not set the files without extensions will be skipped.None
path_spec.table_namestringDisplay name of the dataset.Combination of named variableds from include path and stringsNone
path_spec.enable_compressionbooleanEnable or disable processing compressed files. Currenly .gz and .bz files are supported.True
path_spec.sample_filesbooleanNot listing all the files but only taking a handful amount of sample file to infer the schema. File count and file size calculation will be disabled. This can affect performance significantly if enabledTrue
aws_configAwsConnectionConfig (see below for fields)AWS configuration
aws_config.aws_access_key_idstringAWS access key ID. Can be auto-detected, see for details.None
aws_config.aws_secret_access_keystringAWS secret access key. Can be auto-detected, see for details.None
aws_config.aws_session_tokenstringAWS session token. Can be auto-detected, see for details.None
aws_config.aws_roleGeneric dictAWS roles to assume. If using the string format, the role ARN can be specified directly. If using the object format, the role can be specified in the RoleArn field and additional available arguments are documented at
aws_config.aws_profilestringNamed AWS profile to use. Only used if access key / secret are unset. If not set the default will be usedNone
aws_config.aws_region❓ (required if aws_config is set)stringAWS region code.None
aws_config.aws_endpoint_urlstringAutodetected. See
aws_config.aws_proxyDict[str,string]Autodetected. See
profile_patternsAllowDenyPattern (see below for fields)regex patterns for tables to profile{'allow': ['.*'], 'deny': [], 'ignoreCase': True}
profile_patterns.allowArray of stringList of regex patterns to include in ingestion['.*']
profile_patterns.denyArray of stringList of regex patterns to exclude from ingestion.[]
profile_patterns.ignoreCasebooleanWhether to ignore case sensitivity during pattern matching.True
profilingDataLakeProfilerConfig (see below for fields)Data profiling configuration{'enabled': False, 'profile_table_level_only': False, 'max_number_of_fields_to_profile': None, 'include_field_null_count': True, 'include_field_min_value': True, 'include_field_max_value': True, 'include_field_mean_value': True, 'include_field_median_value': True, 'include_field_stddev_value': True, 'include_field_quantiles': True, 'include_field_distinct_value_frequencies': True, 'include_field_histogram': True, 'include_field_sample_values': True}
profiling.enabledbooleanWhether profiling should be done.False
profiling.profile_table_level_onlybooleanWhether to perform profiling at table-level only or include column-level profiling as well.False
profiling.max_number_of_fields_to_profileintegerA positive integer that specifies the maximum number of columns to profile for any table. None implies all columns. The cost of profiling goes up significantly as the number of columns to profile goes up.None
profiling.include_field_null_countbooleanWhether to profile for the number of nulls for each column.True
profiling.include_field_min_valuebooleanWhether to profile for the min value of numeric columns.True
profiling.include_field_max_valuebooleanWhether to profile for the max value of numeric columns.True
profiling.include_field_mean_valuebooleanWhether to profile for the mean value of numeric columns.True
profiling.include_field_median_valuebooleanWhether to profile for the median value of numeric columns.True
profiling.include_field_stddev_valuebooleanWhether to profile for the standard deviation of numeric columns.True
profiling.include_field_quantilesbooleanWhether to profile for the quantiles of numeric columns.True
profiling.include_field_distinct_value_frequenciesbooleanWhether to profile for distinct value frequencies.True
profiling.include_field_histogrambooleanWhether to profile for the histogram for numeric fields.True
profiling.include_field_sample_valuesbooleanWhether to profile for the sample values for all columns.True

Path Specs

Example - Dataset per file

Bucket structure:

├── employees.csv
└── food_items.csv

Path specs config

- include: s3://test-s3-bucket/*.csv

Example - Datasets with partitions

Bucket structure:

├── orders
│   └── year=2022
│   └── month=2
│   ├── 1.parquet
│   └── 2.parquet
└── returns
└── year=2021
└── month=2
└── 1.parquet

Path specs config:

- include: s3://test-s3-bucket/{table}/{partition_key[0]}={partition[0]}/{partition_key[1]}={partition[1]}/*.parquet

Example - Datasets with partition and exclude

Bucket structure:

├── orders
│   └── year=2022
│   └── month=2
│   ├── 1.parquet
│   └── 2.parquet
└── tmp_orders
└── year=2021
└── month=2
└── 1.parquet

Path specs config:

- include: s3://test-s3-bucket/{table}/{partition_key[0]}={partition[0]}/{partition_key[1]}={partition[1]}/*.parquet
- **/tmp_orders/**

Example - Datasets of mixed nature

Bucket structure:

├── customers
│   ├── part1.json
│   ├── part2.json
│   ├── part3.json
│   └── part4.json
├── employees.csv
├── food_items.csv
├── tmp_10101000.csv
└── orders
   └── year=2022
    └── month=2
   ├── 1.parquet
   ├── 2.parquet
   └── 3.parquet

Path specs config:

- include: s3://test-s3-bucket/*.csv
- **/tmp_10101000.csv
- include: s3://test-s3-bucket/{table}/*.json
- include: s3://test-s3-bucket/{table}/{partition_key[0]}={partition[0]}/{partition_key[1]}={partition[1]}/*.parquet

Valid path_specs.include

s3://my-bucket/foo/tests/bar.avro # single file table   
s3://my-bucket/foo/tests/*.* # mulitple file level tables
s3://my-bucket/foo/tests/{table}/*.avro #table without partition
s3://my-bucket/foo/tests/{table}/*/*.avro #table where partitions are not specified
s3://my-bucket/foo/tests/{table}/*.* # table where no partitions as well as data type specified
s3://my-bucket/{dept}/tests/{table}/*.avro # specifying keywords to be used in display name
s3://my-bucket/{dept}/tests/{table}/{partition_key[0]}={partition[0]}/{partition_key[1]}={partition[1]}/*.avro # specify partition key and value format
s3://my-bucket/{dept}/tests/{table}/{partition[0]}/{partition[1]}/{partition[2]}/*.avro # specify partition value only format
s3://my-bucket/{dept}/tests/{table}/{partition[0]}/{partition[1]}/{partition[2]}/*.* # for all extensions
s3://my-bucket/*/{table}/{partition[0]}/{partition[1]}/{partition[2]}/*.* # table is present at 2 levels down in bucket
s3://my-bucket/*/*/{table}/{partition[0]}/{partition[1]}/{partition[2]}/*.* # table is present at 3 levels down in bucket

Valid path_specs.exclude

  • **/tests/**
  • s3://my-bucket/hr/**
  • */tests/.csv
  • s3://my-bucket/foo/*/my_table/**


  • {table} represents folder for which dataset will be created.
  • include path must end with (. or *.[ext]) to represent leaf level.
  • if *.[ext] is provided then only files with specified type will be scanned.
  • /*/ represents single folder.
  • {partition[i]} represents value of partition.
  • {partition_key[i]} represents name of the partition.
  • While extracting, “i” will be used to match partition_key to partition.
  • all folder levels need to be specified in include. Only exclude path can have ** like matching.
  • exclude path cannot have named variables ( {} ).
  • Folder names should not contain {, }, *, / in their names.
  • {folder} is reserved for internal working. please do not use in named variables.

If you would like to write a more complicated function for resolving file names, then a {transformer} would be a good fit.


Specify as long fixed prefix ( with out /*/ ) as possible in path_specs.include. This will reduce the scanning time and cost, specifically on AWS S3


Running profiling against many tables or over many rows can run up significant costs. While we've done our best to limit the expensiveness of the queries the profiler runs, you should be prudent about the set of tables profiling is enabled on or the frequency of the profiling runs.


If you are ingesting datasets from AWS S3, we recommend running the ingestion on a server in the same region to avoid high egress costs.


Profiles are computed with PyDeequ, which relies on PySpark. Therefore, for computing profiles, we currently require Spark 3.0.3 with Hadoop 3.2 to be installed and the SPARK_HOME and SPARK_VERSION environment variables to be set. The Spark+Hadoop binary can be downloaded here.

For an example guide on setting up PyDeequ on AWS, see this guide.

Code Coordinates

  • Class Name: datahub.ingestion.source.s3.source.S3Source
  • Browse on GitHub


If you've got any questions on configuring ingestion for S3 Data Lake, feel free to ping us on our Slack