Skip to content

Latest commit

 

History

History
530 lines (439 loc) · 29.6 KB

CHANGELOG.md

File metadata and controls

530 lines (439 loc) · 29.6 KB
Raw

Release [1.8.4], 2024-09-17

Improvement

  • The S3 help macro now support a role_arn parameter as an alternative way to provide authentication for S3 based models. Thanks to Mitchell Bregman for the contribution!

Release [1.8.3], 2024-09-01

Bug Fixes

  • An issue was detected when using multiple projections. We solved it and added a test to cover that use case. (#349)

Documentation

  • A CONTRIBUTING.md file was added to the repo. Please follow the instructions prior contributing.

Release [1.8.2], 2024-08-22

New Features

  • ClickHouse projections are now fully supported for table materialization, and partly supported for distributed_table materialization. The projection config should be added to the model config (#342), for instance: 
    {{ config(
       materialized='%s',
       projections=[
           {
               'name': 'your_projection_name',
               'query': 'your_projection_query'
           }
       ]
) }}

Bug Fixes

  • Until this release, when writing tests, it was not possible to pass empty seed data. (#341)
  • When a cluster was used, the adapter left a few __dbt_backup tables in the schema. After the fix, these backup tables are now properly dropped. (#326)
  • Due to this GitHub change, we needed to adjust the docker compose command in our tests. (#334)

Improvements

  • Added a new experimental incremental strategy, insert_overwrite, which replaces existing data in a target table partition by partition. This ensures that only the specified partitions are overwritten, helping to maintain performance and data consistency. Note that this feature has not been tested with distributed models and may not work with such materializations. Anton Bryzgalov Thank you for your contribution! (#201)
  • Schema changes for incremental models were improved and now include a sync_all_column option. Can Bekleyici huge thanks for your contribution. (#332)
  • Previously, view settings were only applied during view creation, not when querying the view. Starting with this release, the settings attribute will now also be applied when running queries on a view. (#324)
  • Enhanced the clickhouse__listagg macro to support single field ordering with optional direction, ensuring compatibility with ClickHouse's sorting limitations. Added validation to prevent the use of multiple order-by fields. (#318)

Documentation

  • Update the docs with the new insert_overwrite incremental strategy. (#331)
  • Add documentation for codec column configuration. (#317)

Release [1.8.1], 2024-07-11

Bug Fix

  • Refresh materialized_view table only if --full-refresh is specified.
  • Fix temporary table creation to support dbt unit tests.
  • ClickHouse dictionaries are now correctly created "on cluster" when a cluster is defined.

Improvements

  • Added database prefix option for local tables of distributed tables (until this change, only a suffix was supported).
  • You can now customize tcp_keepalive configuration for native connections.
  • Implement listagg (groupArray) macro.

Release [1.8.0], 2024-06-13

Improvements

  • Upgrade the connector to use dbt-core 1.8.0. More info about this upgrade can be found here.

Beginning in v1.8, dbt-core and adapters are decoupled. Going forward, your installations should explicitly include both dbt-core and the desired adapter. The new pip installation command should look like this:

pip install dbt-core dbt-clickhouse

Release [1.7.7], 2024-05-31

Bug Fix

  • Fix bool_or behavior (a cross-database dbt macro ).
  • Fix query_settings for models with contracts.
  • Fix the option to use Nullable and LowCardinality in column constraints.
  • Specifying an empty or null value for a connection_overrides field produces invalid DDL for the dictionary materialization. A fix was introduced in this release.

Improvements

Release [1.7.6], 2024-04-12

Bug Fix

  • A bug in (experimental) Distributed Table model creation could lead to errors when there was a change in the model definition (see, e.g., ClickHouse#226). Thanks to Thomas Schmidt for the Fix!
  • A comment at the end of a model would break the query used to retrieve the result column datatypes. Thanks to triou for the bug report and the fix. Closes ClickHouse#256

Improvements

  • The new materialization for ClickHouse dictionaries now takes an optional "credentials dictionary" argument that overrides the global credentials values for user, password, database, host, and port (including removing any of those values by adding empty values if not needed). This allows better control over creating dictionaries on different server. Thanks to Cristhian Garcia for the PR!
  • A new ttl setting has been added to model configuration that will insert the provided ClickHouse TTL expression in the appropriate place. Thanks to Evan Rusackas for the contribution!
  • The Agate library should now be lazy loaded. This should modestly improve dbt startup times (after dbt-clickhouse is upgraded to dbt 1.8.x). Thanks to Daniel Reeves for PR.

Release [1.7.5], 2024-04-02

Bug Fixes

  • Requirements and tests upgraded to include Python 3.12. Closes ClickHouse#264
  • Model settings were not working correctly for custom materializations. Thanks to original dbt-clickhouse silentsokolov for the PR!

Release [1.7.4], 2024-03-23

Improvement

  • Adds support for materializing ClickHouse dictionaries. Thanks to Rory Sawyer for the contribution! See his excellent tests for example usage.

Release [1.7.3], 2024-03-11

Bug Fixes

  • Fixed an issue where passing settings to on view creation didn't work.
  • The dbt test command with a LIMIT clause were broken due to parsing error when having settings in the query (issue). We added a dedicated limit placer, that takes into account the settings section (using a comment flag -- end_of_sql within the query).

Release [1.7.2], 2024-02-09

Bug Fix

  • Fixed an issue where Materialize Views would break with a custom schema. Thanks to Rory Sawyer for the PR!

Release [1.7.1], 2023-12-13

Bug Fixes

  • Some models with LIMIT clauses were broken in recent releases. This has been fixed. Thanks to ptemarvelde for the PR!
  • It was possible for incremental models with the delete+insert strategy to fail if ClickHouse "light weight deletes" were not enabled or the required setting allow_nondetermistic_mutations was not enabled and the user did not have permission to apply it. This condition is now detected on startup, and an exception will be thrown if use_lw_deletes is configured in the profile. Otherwise, a warning will be logged that incremental models will be slower (because such models will be downgraded to use the legacy incremental strategy). This should prevent the confusing behavior in ClickHouse#197 by throwing an early exception for an unsupported configuration.

Release [1.7.0], 2023-12-07

Improvements

  • Minimal compatibility with dbt 1.7.x. The date_spine macro and additional automated tests have not been implemented, but are planned for a future patch release.
  • DBT 1.7 introduces a (complex) optimization mechanism for retrieving a dbt catalog which is overkill for ClickHouse (which has no separate schema/database level), so this release includes some internal catalog changes to simplify that process.

Release [1.6.2], 2023-12-06

Bug Fix

  • The dbt on_schema_change configuration value for incremental models was effectively being ignored. This has been fixed with a very limited implementation. Closes ClickHouse#199. Because of the way that ORDER BY/SORT BY/PARTITION BY/PRIMARY KEYS work in ClickHouse, plus the complexities of correctly transforming ClickHouse data types, sync_all_columns is not currently supported (although an implementation that works for non-key columns is theoretically possible, such an enhancement is not currently planned). Accordingly, only ignore, fail, and append_new_columns values are supported for on_schema_change. It is also not currently supported for Distributed tables.

Note that actually appending new columns requires a fallback to the legacy incremental strategy, which is quite inefficient, so while theoretically possible, using append_new_columns is not recommended except for very small data volumes.

Release [1.6.1], 2023-12-04

Bug Fixes

  • Identifier quoting was disabled for tables/databases etc. This would cause failures for schemas or tables using reserved words or containing special characters. This has been fixed and some macros have been updated to correctly handle such identifiers. Note that there still may be untested edge cases where nonstandard identifiers cause issues, so they are still not recommended. Closes ClickHouse#144. Thanks to Alexandru Pisarenco for the report and initial PR!
  • The new allow_automatic_deduplication setting was not being correctly propagated to the adapter, so setting it to True did not have the intended affect. In addition, this setting is now ignored for older ClickHouse versions that do not support CREATE TABLE AS SELECT ... EMPTY, since the automatic deduplication window is required to allow correct inserts in Replicated tables on those older versions. Fixes ClickHouse#216.

Release [1.6.0], 2023-11-30

Improvements

  • Compatible with dbt 1.6.x. Note that dbt new clone feature is not supported, as ClickHouse has no native "light weight" clone functionality, and copying tables without actual data transfer is not possible in ClickHouse (barring file manipulation outside ClickHouse itself).
  • A new ClickHouse specific Materialized View materialization contributed by Rory Sawyer. This creates a ClickHouse Materialized view using the TO form with the name <model_name>_mv and the associated target table <model_name>. It's highly recommended to fully understand how ClickHouse materialized views work before using this materialization.

Release [1.5.2], 2023-11-28

Bug Fixes

  • The ON CLUSTER clause was in the incorrect place for legacy incremental materializations. This has been fixed. Thanks to Steven Reitsma for the fix!
  • The ON CLUSTER DDL for drop tables did not include a SYNC modifier, which might be the cause of some "table already exists" errors. The SYNC modifier has been added to the on_cluster macro when dropping relations.
  • Fixed a bug where using table settings such as allow_nullable_key would break "legacy" incremental materializations. Closes ClickHouse#209. Also see the new model config property insert_settings described below.
  • Fixed an issue where incremental materializations would incorrectly exclude duplicated inserted elements due to "automatic" ClickHouse deduplication on replicated tables. Closes ClickHouse#213. The fix consists of always sending a replicated_deduplication_window=0 table setting when creating the incremental relations. This behavior can be overridden by setting the new profile parameter allow_automatic_deduplication to True, although for general dbt operations this is probably not necessary and not recommended. Finally thanks to Andy(https://github.com/andy-miracl) for the report and debugging help!

Improvements

  • Added a new profile property allow_automatic_deduplication, which defaults to False. ClickHouse Replicated deduplication is now disable for incremental inserts, but this property can be set to true if for some reason the default ClickHouse behavior for inserted blocks is desired.
  • Added a new model config property query_settings for any ClickHouse settings that should be sent with the INSERT INTO or DELETE_FROM queries used with materializations. Note this is distinct from the existing property settings which is used for ClickHouse "table" settings in DDL statements like CREATE TABLE ... AS.

Release [1.5.1], 2023-11-27

Bug Fix

  • Fix table materialization for compatibility with SQLFluff. Thanks to Kristof Szaloki for the PR!

Release [1.5.0], 2023-11-23

Improvements

  • Compatible with dbt 1.5.x
  • Contract support (using exact column data types)

Bug Fix

  • Fix s3 macro when bucket includes https:// prefix. Closes ClickHouse#192.

Release [1.4.9], 2023-10-27

Improvement

  • Lots of work on Distributed table materializations. Big thanks to gfunc for the additional PR and Zhenbang for code review and suggestions. See the README for details on how to use the new functionality.

Bug Fix

Release [1.4.8], 2023-08-22

Bug Fix

  • Fixed issues with experimental Distributed table materializations. Closes ClickHouse#179. Thanks to Zhebnang for the report and for contributing to the fix with gfunc.

Release [1.4.7], 2023-08-09

Bug Fix

  • Fixed an exception in "legacy" incremental materializations that are not distributed

Release [1.4.6], 2023-07-27

Bug fix

  • Lightweight deletes could fail in environments where the HTTP session was not preserved (such as clusters behind a non-sticky load balancer). This has been fixed by sending the required settings with every request instead of relying on a SET statement. A similar approach has been used to persist the 'insert_distributed_sync' setting for Distributed table materializations.

Release [1.4.5], 2023-07-27

Improvement

  • Adds additional experimental support for Distributed table engine models and incremental materialization. See the README for details. Thanks to gladkikhtutu for the contribution!

Release [1.4.4], 2023-07-19

Bug Fixes

  • Fixed two logging/exception handling issues that would cause exception on startup or when handling some exceptions from the ClickHouse server. Partially addresses ClickHouse#169.
  • Fixed issue with the on_cluster macro that would break the exchange tables step of incremental materializations with an active cluster. Thanks to Andrew Davis for the PR. Closes ClickHouse#167

Release [1.4.3], 2023-06-27

Bug Fix

Improvement

  • Added macros for creating distributed tables. See the distributed_table.sql include file. Thanks to gladkikhtutu for the contribution.

Release [1.4.2], 2023-05-14

Bug fixes

  • Create initial dbt database (if not found) on the defined cluster on first run, instead of just the execution node. Thanks to Jens Hoevenaars for the PR
  • Fix the SYSTEM SYNC REPLICA statement when exchanging tables ON CLUSTER for incremental materializations. Thanks to Schum for PR. Closed ClickHouse#157.

Release [1.4.1], 2023-05-11

Improvements

  • Reduce the number of SQL calls for Modify Comment operations. Thanks to Konstantin Ilchenko.
  • Add "on cluster" to several additional macros to better support distributed tables. Thanks to Saurabh Bikram
  • Add the unique invocation id to temporary "new data" used in delete+insert incremental materializations to allow parallel transformations on the same table. In general parallel transformations are risky, so this approach should only be used when transformations are explicitly limited to non-overlapping data ranges. Closes ClickHouse#150

Release [1.4.0], 2023-02-06

Improvements

  • Support dbt [1.4.1] ClickHouse#135
    • Adds support for Python 3.11
    • Adds additional dbt 1.4.0 tests
    • Adds support for incremental_predicates. This only applies to delete+insert incremental strategy. Note that incremental predicates that depend on "non-deterministic" data (such as a subquery using a table that is accepting inserts) could lead to unexpected results for ReplicatedMergeTree tables depending on the timing of the incremental materialization.
    • Replaces deprecated Exception classes
  • Setting the use_lw_deletes profile value to True will now attempt to enable the allow_experimental_lightweight_delete setting for the dbt session (if user has such permissions on the ClickHouse server). See ClickHouse#133

Bug Fix

  • Composite unique keys specified as a list would not work with incremental materializations. This has been fixed.

Release 1.3.3, 2023-01-18

Documentation Update

  • The documentation has been updated to reflect that dbt-clickhouse does support ephemeral models, and ephemeral model tests do pass. However, due to a ClickHouse limitation, CTEs will not work directly with INSERT statements so table models will fail if they include ephemeral models in the SELECT. View models and other SQL statements using ephemeral models should work correctly.

Bug Fix

  • Client connections would incorrectly reuse a previous session_id when initializing, causing session locked errors. This has been fixed. This closes ClickHouse#127. Multiple threads should now work correctly in dbt projects, and dbt-clickhouse automated tests now use threads=4 (soon to be the dbt default).

Release 1.3.2, 2022-12-23

Improvements

  • Added experimental support for the delete+insert incremental strategy. In most cases this strategy should be significantly faster than the existing "legacy" custom strategy (which is still the default). To enable delete+insert incremental materialization, the flag allow_experimental_lightweight_delete must be enabled on your ClickHouse server. This flag is NOT currently considered production ready, so use at your own risk. To use this strategy as the "default" incremental strategy, the new configuration value use_lw_deletes must be set to True. Otherwise, to use it for a particular model, set the model's incremental_strategy to delete+insert. Important caveats about this strategy:
    • This strategy directly uses lightweight deletes on the target model/table. It does not create a temporary or intermediate table. Accordingly, if there is an issue during that transformation, the materialization can not be rolled back and the model may contain inconsistent data.
    • If the incremental materialization includes schema changes, or lightweight deletes are not available, dbt-clickhouse will fall back to the much slower "legacy" strategy.
  • Allow append as an incremental strategy. This is the same as using the custom model configuration value inserts_only.
  • New s3source macro. This simplifies the process of using the ClickHouse s3 table function in queries. See the tests for example usage.

Bug Fixes

  • The ON CLUSTER clause has been added to additional DDL statements including incremental models processing. Closes ClickHouse#117 and should close ClickHouse#95 for Replicated tables that use the {uuid} macro in the path to avoid name conflicts. Thanks to Saurabh Bikram
  • The apply and revoke grants macros now correctly work with roles as well as user. Again thanks to Saurabh Bikram
  • A compound unique_key (such as key1, key2) now works correctly with incremental models

Release 1.3.1, 2022-11-17

Improvements

  • Improve error message when atomic "EXCHANGE TABLES" operation is not supported

Release 1.3.0, 2022-10-30

Improvements

  • Support dbt 1.3.0 ClickHouse#105
    • Adds additional dbt 1.3.0 core tests
    • Adds array utility macros ported from dbt-utils
    • Does NOT add support for Python models (not implemented)
    • Does NOT utilize default/standard incremental materialization macros (standard strategies do not work in ClickHouse)

Bug Fix

  • Require exact match for relations. ClickHouse databases and tables are all case-sensitive, so all searches are now case-sensitive. Closes ClickHouse#100 and ClickHouse#110



Release 1.2.1, 2022-09-19

Improvements

  • Support dbt 1.2.1 ClickHouse#79
    • Grants support on models
    • Support the cross database dbt-core macros (migrated from dbt-utils)
    • Add docs tests
    • Validate Python 3.10 support
    • Implement retry logic for small set of "retryable" connection errors
  • Don't close connection on release to improvement model performance
  • Allow specifying database engine for schema creation
  • Extend support for ClickHouse Cloud and Replicated databases

Bug Fix

  • Validate that atomic EXCHANGE TABLES is supported on the target ClickHouse server ClickHouse#94



Release 1.1.8, 2022-09-02

Improvement

  • Support use of atomic EXCHANGE TABLES when supported by ClickHouse



Release 1.1.7, 2022-07-11

Improvement

  • Support ClickHouse Arrays with Nullable value types

Bug Fix

  • Fix brackets regex in columns)

Release 1.1.6, 2022-07-04

Improvement

  • Add support for CREATE AS SELECT in Replicated Databases



Release 1.1.5, 2022-06-27

Bug Fixes

  • Ensure database exists before connection
  • Safely handle dropping working database



Release 1.1.4, 2022-06-27

Improvement

  • Allow selection of either native protocol driver (clickhouse-driver) or http driver (clickhouse-connect)



Release 1.1.2, 2022-06-22

Rename from v 1.1.0.2



Release 1.1.0.2, 2022-06-17

Improvements

  • Support for inserting SETTINGS section to CREATE AS and INSERT INTO queries through model configuration
  • Support adding custom session settings to the connection through profile credentials.
  • Allow using custom databases/schemas for connection
  • Support Boolean fields in CSV seeds

Bug Fix

  • Fixed prehooks and empty fields in CSV seeds



Release 1.1.0, 2022-06-02

Release 1.1.0.1, 2022-06-09

Bug Fix

  • Fixed prehooks and empty fields in CSV seeds



Release 1.1.0, 2022-06-02

Improvements

  • 1.1.0 dbt support
  • Snapshots timestamps
  • Replace temporary in-memory tables with MergeTree table - removing memory limitations over Incremental model and snapshots creation
  • Moved to use clickhouse-connect driver - an officially supported HTTP driver by ClickHouse Inc.

Release 1.0.4, 2022-04-02

Improvements

  • 1.0.4 dbt support
  • New logger

Release 1.0.1 - 2022-02-09

Improvement

  • Support 1.0.1 dbt

Bug Fixes

Release 1.0.0, 2022-01-02

Add

  • dbt 1.0.0 support

Release 0.21.1, 2022-01-01

Improvements

  • dbt 0.21.1 support
  • Extended settings for clickhouse-driver ClickHouse#27

Bug Fixes

Release 0.21.0, 2021-11-18

Improvement

  • Support 0.21.0 dbt

Bug Fixes

Release 0.20.2, 2021-10-16

Improvement

  • dbt 0.20.1 support

Improvement

Bug Fixes

Release 0.20.1, 2021-08-15

Improvement

  • dbt 0.20.1 support

Release 0.20.0, 2021-08-14

Improvement

  • dbt 0.20.0 support

Release 0.19.1.1, 2021-08-13

Improvement

  • Add verify and secure to connection configuration

Bug Fix

Release 0.19.1, 2021-05-07

Improvement

  • Add support the ON CLUSTER clause for main cases

Usage Change

  • Engine now require brackets ()

Bug Fix

  • Fix a missing sample profile

Release 0.19.0.2, 2021-04-03

Bug Fix

  • Fix name of partition

Release 0.19.0.1, 2021-03-30

Bug Fix

  • Fix version of clickhouse-driver in setup.py

Release 0.19.0, 2021-02-14

Initial Release