hide:
There are three ways to pass in configuration:
.pyiceberg.yaml configuration file (Recommended)The configuration file can be stored in either the directory specified by the PYICEBERG_HOME environment variable, the home directory, or current working directory (in this order).
To change the path searched for the .pyiceberg.yaml, you can overwrite the PYICEBERG_HOME environment variable.
Another option is through environment variables:
export PYICEBERG_CATALOG__DEFAULT__URI=thrift://localhost:9083 export PYICEBERG_CATALOG__DEFAULT__S3__ACCESS_KEY_ID=username export PYICEBERG_CATALOG__DEFAULT__S3__SECRET_ACCESS_KEY=password
The environment variable picked up by Iceberg starts with PYICEBERG_ and then follows the yaml structure below, where a double underscore __ represents a nested field, and the underscore _ is converted into a dash -.
For example, PYICEBERG_CATALOG__DEFAULT__S3__ACCESS_KEY_ID, sets s3.access-key-id on the default catalog.
Iceberg tables support table properties to configure table behavior.
| Key | Options | Default | Description |
|---|---|---|---|
write.parquet.compression-codec | {uncompressed,zstd,gzip,snappy} | zstd | Sets the Parquet compression coddec. |
write.parquet.compression-level | Integer | null | Parquet compression level for the codec. If not set, it is up to PyIceberg |
write.parquet.row-group-limit | Number of rows | 1048576 | The upper bound of the number of entries within a single row group |
write.parquet.page-size-bytes | Size in bytes | 1MB | Set a target threshold for the approximate encoded size of data pages within a column chunk |
write.parquet.page-row-limit | Number of rows | 20000 | Set a target threshold for the maximum number of rows within a column chunk |
write.parquet.dict-size-bytes | Size in bytes | 2MB | Set the dictionary page size limit per row group |
write.metadata.previous-versions-max | Integer | 100 | The max number of previous version metadata files to keep before deleting after commit. |
write.metadata.delete-after-commit.enabled | Boolean | False | Whether to automatically delete old tracked metadata files after each table commit. It will retain a number of the most recent metadata files, which can be set using property write.metadata.previous-versions-max. |
write.object-storage.enabled | Boolean | True | Enables the ObjectStoreLocationProvider that adds a hash component to file paths. Note: the default value of True differs from Iceberg's Java implementation |
write.object-storage.partitioned-paths | Boolean | True | Controls whether partition values are included in file paths when object storage is enabled |
write.py-location-provider.impl | String of form module.ClassName | null | Optional, custom LocationProvider implementation |
write.data.path | String pointing to location | {metadata.location}/data | Sets the location under which data is written. |
write.metadata.path | String pointing to location | {metadata.location}/metadata | Sets the location under which metadata is written. |
| Key | Options | Default | Description |
|---|---|---|---|
commit.manifest.target-size-bytes | Size in bytes | 8388608 (8MB) | Target size when merging manifest files |
commit.manifest.min-count-to-merge | Number of manifests | 100 | Target size when merging manifest files |
commit.manifest-merge.enabled | Boolean | False | Controls whether to automatically merge manifests on writes |
!!! note “Fast append” Unlike Java implementation, PyIceberg default to the fast append and thus commit.manifest-merge.enabled is set to False by default.
Iceberg works with the concept of a FileIO which is a pluggable module for reading, writing, and deleting files. By default, PyIceberg will try to initialize the FileIO that‘s suitable for the scheme (s3://, gs://, etc.) and will use the first one that’s installed.
PyArrowFileIO, FsspecFileIOPyArrowFileIOPyArrowFileIOPyArrowFileIOFsspecFileIOPyArrowFileIOYou can also set the FileIO explicitly:
| Key | Example | Description |
|---|---|---|
| py-io-impl | pyiceberg.io.fsspec.FsspecFileIO | Sets the FileIO explicitly to an implementation, and will fail explicitly if it can't be loaded |
For the FileIO there are several configuration options available:
| Key | Example | Description |
|---|---|---|
| s3.endpoint | https://10.0.19.25/ | Configure an alternative endpoint of the S3 service for the FileIO to access. This could be used to use S3FileIO with any s3-compatible object storage service that has a different endpoint, or access a private S3 endpoint in a virtual private cloud. |
| s3.access-key-id | admin | Configure the static access key id used to access the FileIO. |
| s3.secret-access-key | password | Configure the static secret access key used to access the FileIO. |
| s3.session-token | AQoDYXdzEJr... | Configure the static session token used to access the FileIO. |
| s3.role-session-name | session | An optional identifier for the assumed role session. |
| s3.role-arn | arn:aws:... | AWS Role ARN. If provided instead of access_key and secret_key, temporary credentials will be fetched by assuming this role. |
| s3.signer | bearer | Configure the signature version of the FileIO. |
| s3.signer.uri | http://my.signer:8080/s3 | Configure the remote signing uri if it differs from the catalog uri. Remote signing is only implemented for FsspecFileIO. The final request is sent to <s3.signer.uri>/<s3.signer.endpoint>. |
| s3.signer.endpoint | v1/main/s3-sign | Configure the remote signing endpoint. Remote signing is only implemented for FsspecFileIO. The final request is sent to <s3.signer.uri>/<s3.signer.endpoint>. (default : v1/aws/s3/sign). |
| s3.region | us-west-2 | Configure the default region used to initialize an S3FileSystem. PyArrowFileIO attempts to automatically tries to resolve the region if this isn't set (only supported for AWS S3 Buckets). |
| s3.resolve-region | False | Only supported for PyArrowFileIO, when enabled, it will always try to resolve the location of the bucket (only supported for AWS S3 Buckets). |
| s3.proxy-uri | http://my.proxy.com:8080 | Configure the proxy server to be used by the FileIO. |
| s3.connect-timeout | 60.0 | Configure socket connection timeout, in seconds. |
| s3.request-timeout | 60.0 | Configure socket read timeouts on Windows and macOS, in seconds. |
| s3.force-virtual-addressing | False | Whether to use virtual addressing of buckets. If true, then virtual addressing is always enabled. If false, then virtual addressing is only enabled if endpoint_override is empty. This can be used for non-AWS backends that only support virtual hosted-style access. |
| Key | Example | Description |
|---|---|---|
| hdfs.host | https://10.0.19.25/ | Configure the HDFS host to connect to |
| hdfs.port | 9000 | Configure the HDFS port to connect to. |
| hdfs.user | user | Configure the HDFS username used for connection. |
| hdfs.kerberos_ticket | kerberos_ticket | Configure the path to the Kerberos ticket cache. |
| Key | Example | Description |
|---|---|---|
| adls.connection-string | AccountName=devstoreaccount1;AccountKey=Eby8vdM02xNOcqF...;BlobEndpoint=http://localhost/ | A connection string. This could be used to use FileIO with any adls-compatible object storage service that has a different endpoint (like azurite). |
| adls.account-name | devstoreaccount1 | The account that you want to connect to |
| adls.account-key | Eby8vdM02xNOcqF... | The key to authentication against the account. |
| adls.sas-token | NuHOuuzdQN7VRM%2FOpOeqBlawRCA845IY05h9eu1Yte4%3D | The shared access signature |
| adls.tenant-id | ad667be4-b811-11ed-afa1-0242ac120002 | The tenant-id |
| adls.client-id | ad667be4-b811-11ed-afa1-0242ac120002 | The client-id |
| adls.client-secret | oCA3R6P*ka#oa1Sms2J74z... | The client-secret |
| Key | Example | Description |
|---|---|---|
| gcs.project-id | my-gcp-project | Configure Google Cloud Project for GCS FileIO. |
| gcs.oauth2.token | ya29.dr.AfM... | String representation of the access token used for temporary access. |
| gcs.oauth2.token-expires-at | 1690971805918 | Configure expiration for credential generated with an access token. Milliseconds since epoch |
| gcs.access | read_only | Configure client to have specific access. Must be one of ‘read_only’, ‘read_write’, or ‘full_control’ |
| gcs.consistency | md5 | Configure the check method when writing files. Must be one of ‘none’, ‘size’, or ‘md5’ |
| gcs.cache-timeout | 60 | Configure the cache expiration time in seconds for object metadata cache |
| gcs.requester-pays | False | Configure whether to use requester-pays requests |
| gcs.session-kwargs | {} | Configure a dict of parameters to pass on to aiohttp.ClientSession; can contain, for example, proxy settings. |
| gcs.service.host | http://0.0.0.0:4443 | Configure an alternative endpoint for the GCS FileIO to access (format protocol://host:port) If not given, defaults to the value of environment variable “STORAGE_EMULATOR_HOST”; if that is not set either, will use the standard Google endpoint. |
| gcs.default-location | US | Configure the default location where buckets are created, like ‘US’ or ‘EUROPE-WEST3’. |
| gcs.version-aware | False | Configure whether to support object versioning on the GCS bucket. |
PyIceberg uses S3FileSystem class to connect to OSS bucket as the service is compatible with S3 SDK as long as the endpoint is addressed with virtual hosted style.
| Key | Example | Description |
|---|---|---|
| s3.endpoint | https://s3.oss-your-bucket-region.aliyuncs.com/ | Configure an endpoint of the OSS service for the FileIO to access. Be sure to use S3 compatible endpoint as given in the example. |
| s3.access-key-id | admin | Configure the static access key id used to access the FileIO. |
| s3.secret-access-key | password | Configure the static secret access key used to access the FileIO. |
| s3.session-token | AQoDYXdzEJr... | Configure the static session token used to access the FileIO. |
| s3.force-virtual-addressing | True | Whether to use virtual addressing of buckets. This must be set to True as OSS can only be accessed with virtual hosted style address. |
| Key | Example | Description |
|---|---|---|
| pyarrow.use-large-types-on-read | True | Use large PyArrow types i.e. large_string, large_binary and large_list field types on table scans. The default value is True. |
Apache Iceberg uses the concept of a LocationProvider to manage file paths for a table's data files. In PyIceberg, the LocationProvider module is designed to be pluggable, allowing customization for specific use cases, and to additionally determine metadata file locations. The LocationProvider for a table can be specified through table properties.
Both data file and metadata file locations can be customized by configuring the table properties write.data.path and write.metadata.path, respectively.
For more granular control, you can override the LocationProvider's new_data_location and new_metadata_location methods to define custom logic for generating file paths. See Loading a Custom Location Provider.
PyIceberg defaults to the ObjectStoreLocationProvider, which generates file paths for data files that are optimized for object storage.
The SimpleLocationProvider provides paths prefixed by {location}/data/, where location comes from the table metadata. This can be overridden by setting write.data.path table configuration.
For example, a non-partitioned table might have a data file with location:
s3://bucket/ns/table/data/0000-0-5affc076-96a4-48f2-9cd2-d5efbc9f0c94-00001.parquet
When the table is partitioned, files under a given partition are grouped into a subdirectory, with that partition key and value as the directory name - this is known as the Hive-style partition path format. For example, a table partitioned over a string column category might have a data file with location:
s3://bucket/ns/table/data/category=orders/0000-0-5affc076-96a4-48f2-9cd2-d5efbc9f0c94-00001.parquet
The SimpleLocationProvider is enabled for a table by explicitly setting its write.object-storage.enabled table property to False.
PyIceberg offers the ObjectStoreLocationProvider, and an optional partition-exclusion optimization, designed for tables stored in object storage. For additional context and motivation concerning these configurations, see their documentation for Iceberg's Java implementation.
When several files are stored under the same prefix, cloud object stores such as S3 often throttle requests on prefixes, resulting in slowdowns. The ObjectStoreLocationProvider counteracts this by injecting deterministic hashes, in the form of binary directories, into file paths, to distribute files across a larger number of object store prefixes.
Paths are prefixed by {location}/data/, where location comes from the table metadata, in a similar manner to the SimpleLocationProvider. This can be overridden by setting write.data.path table configuration.
For example, a table partitioned over a string column category might have a data file with location: (note the additional binary directories)
s3://bucket/ns/table/data/0101/0110/1001/10110010/category=orders/0000-0-5affc076-96a4-48f2-9cd2-d5efbc9f0c94-00001.parquet
The write.object-storage.enabled table property determines whether the ObjectStoreLocationProvider is enabled for a table. It is used by default.
When the ObjectStoreLocationProvider is used, the table property write.object-storage.partitioned-paths, which defaults to True, can be set to False as an additional optimization for object stores. This omits partition keys and values from data file paths entirely to further reduce key size. With it disabled, the same data file above would instead be written to: (note the absence of category=orders)
s3://bucket/ns/table/data/1101/0100/1011/00111010-00000-0-5affc076-96a4-48f2-9cd2-d5efbc9f0c94-00001.parquet
Similar to FileIO, a custom LocationProvider may be provided for a table by concretely subclassing the abstract base class LocationProvider.
The table property write.py-location-provider.impl should be set to the fully-qualified name of the custom LocationProvider (i.e. mymodule.MyLocationProvider). Recall that a LocationProvider is configured per-table, permitting different location provision for different tables. Note also that Iceberg's Java implementation uses a different table property, write.location-provider.impl, for custom Java implementations.
An example, custom LocationProvider implementation is shown below.
import uuid class UUIDLocationProvider(LocationProvider): def __init__(self, table_location: str, table_properties: Properties): super().__init__(table_location, table_properties) def new_data_location(self, data_file_name: str, partition_key: Optional[PartitionKey] = None) -> str: # Can use any custom method to generate a file path given the partitioning information and file name prefix = f"{self.table_location}/{uuid.uuid4()}" return f"{prefix}/{partition_key.to_path()}/{data_file_name}" if partition_key else f"{prefix}/{data_file_name}"
PyIceberg currently has native catalog type support for REST, SQL, Hive, Glue and DynamoDB. Alternatively, you can also directly set the catalog implementation:
| Key | Example | Description |
|---|---|---|
| type | rest | Type of catalog, one of rest, sql, hive, glue, dymamodb. Default to rest |
| py-catalog-impl | mypackage.mymodule.MyCatalog | Sets the catalog explicitly to an implementation, and will fail explicitly if it can't be loaded |
catalog: default: uri: http://rest-catalog/ws/ credential: t-1234:secret default-mtls-secured-catalog: uri: https://rest-catalog/ws/ ssl: client: cert: /absolute/path/to/client.crt key: /absolute/path/to/client.key cabundle: /absolute/path/to/cabundle.pem
| Key | Example | Description |
|---|---|---|
| uri | https://rest-catalog/ws | URI identifying the REST Server |
| ugi | t-1234:secret | Hadoop UGI for Hive client. |
| credential | t-1234:secret | Credential to use for OAuth2 credential flow when initializing the catalog |
| token | FEW23.DFSDF.FSDF | Bearer token value to use for Authorization header |
| scope | openid offline corpds:ds:profile | Desired scope of the requested security token (default : catalog) |
| resource | rest_catalog.iceberg.com | URI for the target resource or service |
| audience | rest_catalog | Logical name of target resource or service |
| rest.sigv4-enabled | true | Sign requests to the REST Server using AWS SigV4 protocol |
| rest.signing-region | us-east-1 | The region to use when SigV4 signing a request |
| rest.signing-name | execute-api | The service signing name to use when SigV4 signing a request |
| oauth2-server-uri | https://auth-service/cc | Authentication URL to use for client credentials authentication (default: uri + ‘v1/oauth/tokens’) |
To configure custom headers in RESTCatalog, include them in the catalog properties with the prefix header.. This ensures that all HTTP requests to the REST service include the specified headers.
catalog: default: uri: http://rest-catalog/ws/ credential: t-1234:secret header.content-type: application/vnd.api+json
Specific headers defined by the RESTCatalog spec include:
| Key | Options | Default | Description |
|---|---|---|---|
header.X-Iceberg-Access-Delegation | {vended-credentials,remote-signing} | vended-credentials | Signal to the server that the client supports delegated access via a comma-separated list of access mechanisms. The server may choose to supply access via any or none of the requested mechanisms |
The SQL catalog requires a database for its backend. PyIceberg supports PostgreSQL and SQLite through psycopg2. The database connection has to be configured using the uri property. The init_catalog_tables is optional and defaults to True. If it is set to False, the catalog tables will not be created when the SQLCatalog is initialized. See SQLAlchemy's documentation for URL format:
For PostgreSQL:
catalog: default: type: sql uri: postgresql+psycopg2://username:password@localhost/mydatabase init_catalog_tables: false
In the case of SQLite:
!!! warning inline end “Development only” SQLite is not built for concurrency, you should use this catalog for exploratory or development purposes.
catalog: default: type: sql uri: sqlite:////tmp/pyiceberg.db init_catalog_tables: false
| Key | Example | Default | Description |
|---|---|---|---|
| uri | postgresql+psycopg2://username:password@localhost/mydatabase | SQLAlchemy backend URL for the catalog database (see documentation for URL format) | |
| echo | true | false | SQLAlchemy engine echo param to log all statements to the default log handler |
| pool_pre_ping | true | false | SQLAlchemy engine pool_pre_ping param to test connections for liveness upon each checkout |
The in-memory catalog is built on top of SqlCatalog and uses SQLite in-memory database for its backend.
It is useful for test, demo, and playground but not in production as it does not support concurrent access.
catalog: default: type: in-memory warehouse: /tmp/pyiceberg/warehouse
| Key | Example | Default | Description |
|---|---|---|---|
| warehouse | /tmp/pyiceberg/warehouse | file:///tmp/iceberg/warehouse | The directory where the in-memory catalog will store its data files. |
catalog: default: uri: thrift://localhost:9083 s3.endpoint: http://localhost:9000 s3.access-key-id: admin s3.secret-access-key: password
| Key | Example | Description |
|---|---|---|
| hive.hive2-compatible | true | Using Hive 2.x compatibility mode |
| hive.kerberos-authentication | true | Using authentication via Kerberos |
When using Hive 2.x, make sure to set the compatibility flag:
catalog: default: ... hive.hive2-compatible: true
Your AWS credentials can be passed directly through the Python API. Otherwise, please refer to How to configure AWS credentials to set your AWS account credentials locally.
catalog: default: type: glue glue.access-key-id: <ACCESS_KEY_ID> glue.secret-access-key: <SECRET_ACCESS_KEY> glue.session-token: <SESSION_TOKEN> glue.region: <REGION_NAME> s3.endpoint: http://localhost:9000 s3.access-key-id: admin s3.secret-access-key: password
catalog: default: type: glue glue.profile-name: <PROFILE_NAME> glue.region: <REGION_NAME> s3.endpoint: http://localhost:9000 s3.access-key-id: admin s3.secret-access-key: password
!!! Note “Client-specific Properties” glue.* properties are for Glue Catalog only. If you want to use the same credentials for both Glue Catalog and S3 FileIO, you can set the client.* properties. See the Unified AWS Credentials section for more details.
| Key | Example | Description |
|---|---|---|
| glue.id | 111111111111 | Configure the 12-digit ID of the Glue Catalog |
| glue.skip-archive | true | Configure whether to skip the archival of older table versions. Default to true |
| glue.endpoint | https://glue.us-east-1.amazonaws.com | Configure an alternative endpoint of the Glue service for GlueCatalog to access |
| glue.profile-name | default | Configure the static profile used to access the Glue Catalog |
| glue.region | us-east-1 | Set the region of the Glue Catalog |
| glue.access-key-id | admin | Configure the static access key id used to access the Glue Catalog |
| glue.secret-access-key | password | Configure the static secret access key used to access the Glue Catalog |
| glue.session-token | AQoDYXdzEJr... | Configure the static session token used to access the Glue Catalog |
| glue.max-retries | 10 | Configure the maximum number of retries for the Glue service calls |
| glue.retry-mode | standard | Configure the retry mode for the Glue service. Default to standard. |
!!! warning “Removed Properties” The properties profile_name, region_name, aws_access_key_id, aws_secret_access_key, and aws_session_token were deprecated and removed in 0.8.0
If you want to use AWS DynamoDB as the catalog, you can use the last two ways to configure the pyiceberg and refer How to configure AWS credentials to set your AWS account credentials locally. If you want to use the same credentials for both Dynamodb Catalog and S3 FileIO, you can set the client.* properties.
catalog: default: type: dynamodb table-name: iceberg
If you prefer to pass the credentials explicitly to the client instead of relying on environment variables,
catalog: default: type: dynamodb table-name: iceberg dynamodb.access-key-id: <ACCESS_KEY_ID> dynamodb.secret-access-key: <SECRET_ACCESS_KEY> dynamodb.session-token: <SESSION_TOKEN> dynamodb.region: <REGION_NAME> s3.endpoint: http://localhost:9000 s3.access-key-id: admin s3.secret-access-key: password
!!! Note “Client-specific Properties” dynamodb.* properties are for DynamoDB Catalog only. If you want to use the same credentials for both DynamoDB Catalog and S3 FileIO, you can set the client.* properties. See the Unified AWS Credentials section for more details.
| Key | Example | Description |
|---|---|---|
| dynamodb.profile-name | default | Configure the static profile used to access the DynamoDB Catalog |
| dynamodb.region | us-east-1 | Set the region of the DynamoDB Catalog |
| dynamodb.access-key-id | admin | Configure the static access key id used to access the DynamoDB Catalog |
| dynamodb.secret-access-key | password | Configure the static secret access key used to access the DynamoDB Catalog |
| dynamodb.session-token | AQoDYXdzEJr... | Configure the static session token used to access the DynamoDB Catalog |
!!! warning “Removed Properties” The properties profile_name, region_name, aws_access_key_id, aws_secret_access_key, and aws_session_token were deprecated and removed in 0.8.0
If you want to load any custom catalog implementation, you can set catalog configurations like the following:
catalog: default: py-catalog-impl: mypackage.mymodule.MyCatalog custom-key1: value1 custom-key2: value2
You can explicitly set the AWS credentials for both Glue/DynamoDB Catalog and S3 FileIO by configuring client.* properties. For example:
catalog: default: type: glue client.access-key-id: <ACCESS_KEY_ID> client.secret-access-key: <SECRET_ACCESS_KEY> client.region: <REGION_NAME>
configures the AWS credentials for both Glue Catalog and S3 FileIO.
| Key | Example | Description |
|---|---|---|
| client.region | us-east-1 | Set the region of both the Glue/DynamoDB Catalog and the S3 FileIO |
| client.access-key-id | admin | Configure the static access key id used to access both the Glue/DynamoDB Catalog and the S3 FileIO |
| client.secret-access-key | password | Configure the static secret access key used to access both the Glue/DynamoDB Catalog and the S3 FileIO |
| client.session-token | AQoDYXdzEJr... | Configure the static session token used to access both the Glue/DynamoDB Catalog and the S3 FileIO |
| client.role-session-name | session | An optional identifier for the assumed role session. |
| client.role-arn | arn:aws:... | AWS Role ARN. If provided instead of access_key and secret_key, temporary credentials will be fetched by assuming this role. |
!!! Note “Properties Priority” client.* properties will be overridden by service-specific properties if they are set. For example, if client.region is set to us-west-1 and s3.region is set to us-east-1, the S3 FileIO will use us-east-1 as the region.
PyIceberg uses multiple threads to parallelize operations. The number of workers can be configured by supplying a max-workers entry in the configuration file, or by setting the PYICEBERG_MAX_WORKERS environment variable. The default value depends on the system hardware and Python version. See the Python documentation for more details.
Previous versions of Java (<1.4.0) implementations incorrectly assume the optional attribute current-snapshot-id to be a required attribute in TableMetadata. This means that if current-snapshot-id is missing in the metadata file (e.g. on table creation), the application will throw an exception without being able to load the table. This assumption has been corrected in more recent Iceberg versions. However, it is possible to force PyIceberg to create a table with a metadata file that will be compatible with previous versions. This can be configured by setting the legacy-current-snapshot-id property as “True” in the configuration file, or by setting the PYICEBERG_LEGACY_CURRENT_SNAPSHOT_ID environment variable. Refer to the PR discussion for more details on the issue
PyIceberg currently only supports upto microsecond precision in its TimestampType. PyArrow timestamp types in ‘s’ and ‘ms’ will be upcast automatically to ‘us’ precision timestamps on write. Timestamps in ‘ns’ precision can also be downcast automatically on write if desired. This can be configured by setting the downcast-ns-timestamp-to-us-on-write property as “True” in the configuration file, or by setting the PYICEBERG_DOWNCAST_NS_TIMESTAMP_TO_US_ON_WRITE environment variable. Refer to the nanoseconds timestamp proposal document for more details on the long term roadmap for nanoseconds support