Version:

Changelog

Changelog for all the Profiles versions.

58 minute read
Last modified: Nov 14, 2024

Version 0.19

14 November 2024
Schema version: 84

What’s New

Schema has been updated from 80 to 84.
Project created using pb init pb-project now locks the PB version.
Task summary is now shown at the end of each run, including relevant command details like start time, end time, sequence number, and total models processed (if any).
The number of iterations/loops during ID stitching are now logged.
A new feature named Profiles Copilot is introduced. It is embedded within the PB core CLI tool and helps you analyze the health of your ID graph by giving you a high level aggregate analysis of your ID graph across entities. You can also analyze a single cluster by visualizing all IDs and their connections, along with highlighting the most important IDs that form connections in that group.
A key named carry_forward_privileges (type Boolean) is added in the pb_project.yaml file. This key determines whether the privileges granted on the views should be retained after the view definitions are updated. This resolves the issue where recreating views for each model in the project led to revoked privileges. Note that using this key can lead to some performance overhead.
A new column named pb_version is added to the material_registry table in the database which denotes the PB version that created the entry.
A new flag pb show idstitcher-report --seed_clusters_csv is added which you can use to pass a CSV file with two columns - id and id_type. The cluster ID can be either rudder_id or other_id. Each row specifies an ID to be mapped to a cluster main ID, which is then included in the report.

Improvements

An error is now thrown if a model has a feature without any IDs.
User-friendly messages are shown in case a package isn’t installed.
Error is now thrown if duplicate feature names are present for a cohort.

Bug fixes

Resolved the cleanup issue where some materials cleanup failed due to dependencies.
Fixed an issue where a non-existent folder inside model_folders caused the project to fail.

Known Issues

BigQuery

pb validate access command does not work for BigQuery.

Redshift

If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
Cross database references can fail on Redshift for a few clusters.
While creating Activations, validation for Redshift does not work correctly in the RudderStack dashboard.

Other issues

Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
pb insert does not work for Redshift, Databricks, and BigQuery.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
Timegrains is an experimental feature. There might be some undiscovered issues.

Version 0.18.4

25 October 2024

Bug Fixes

The schema name has been prepended to the drop statements executed during cleanup. This ensures that deletions are always performed in the correct schema.

Version 0.18.3

16 October 2024

Bug Fixes

Fixed a bug that causes cleanup of materials fail due to current transaction is aborted error. With the fix, if cleanup of one material fails (for some reason, ex: other objects depend on it), the cleanup of other expired materials should continue.

Version 0.18.2

4 October 2024

Bug Fixes

Resolved a migration bug which occurs when there are nested model folders containing non yaml files.
Cleanup with flag --remove_latest_view_ptrs was not respecting retention time period set in pb_project.yaml. This is fixed.

Version 0.18.1

3 October 2024

Bug Fixes

Fixed issue in default ID stitcher for run_type: discrete.
Resolved a bug that occurs when a project contains multiple entities with same cohort name.
During migration, pb was skipping non-YAML files, which caused scheduled runs to fail. This is fixed.

Version 0.18

27 September 2024
Schema version: 80

What’s New

Cohort model now lets you perform filtering using a filter_expression followed by AND/OR list of expressions, for example:

models:
   - name: high_value_us_residents
     model_type: cohort
     model_spec:
       ...
       filter_expression:
         AND:
           - {{ user.Var('country') }} = 'US'
           - {{ user.Var('salary') }} > 10000

You can define the retention_period for each model of a project. Further, the pb cleanup materials --expired command cleans up the materials beyond the defined retention period.
Referring other entity_vars/input_vars is now simplified. You can use {{entityName.entity_varName}} instead of the earlier one {{entityName.Var("entity_varName")}}. Note that the earlier syntax also works fine.
You can use features of an SQL model while using a cohort. To do so, specify the entity_key or entity_cohort in the model_spec of an SQL model.
pb cleanup materials --concurrency - A new command which enables concurrency for cleanup, by defining the number of concurrent workers for cleanup. The default value is 1.
The default offset value while executing pb run command is now updated to 0. It was 30 minutes earlier.
A new flag --end_time_offset is added to the compile/run commands for adding an offset to the end timestamp, in a human readable format. It means that RudderStack does not use any data you load in the warehouse after the offset time has elapsed for that run. For example, pb run --end_time_offset=45m ensures that RudderStack does not use any data older than 45 minutes from the run’s start time. Note that you can’t use this new flag with the seq_no or end_time flags.
You can now import Packages starting with SSH URLs, for example, ssh://git@host:port/path.git.
You can run or import projects hosted on S3 as packages by adding block_store_creds in your site configuration file. To run the project, execute pb run -p s3://<url> command.
Running a project with the --migrate_on_load flag now stores generated artifacts in the output subfolder instead of migrations.
For an entity_var/input_var, the default key has been renamed to default_value.
Simplified the project created using pb init pb-project by removing the dependency on corelib package , sample SQL model, model contracts and CSV’s in the inputs file.
RudderStack now uses INNER JOIN instead of RIGHT JOIN when calculating entity_vars. This results in performance improvement and also prevents some values from getting lost.
Feature view model with main_id as an identifier is created by default.
Schema has been updated from 72 to 80.

Improvements

By default, RudderStack ignores all the blank values in the ID stitcher model.
There is a slight aesthetic improvement in HTML reports generated using pb show idstitcher-report command.
Relevant errors are now thrown if you specify an unknown YAML key in the model definition.

Bug Fixes

validity_time key has been removed.
The pb validate access command, for Databricks, now checks only for the necessary permissions and not for ALL the privileges.

Known Issues

BigQuery

pb validate access command does not work for BigQuery.

Redshift

If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
Cross database references can fail on Redshift for a few clusters.
While creating Activations, validation for Redshift does not work correctly in the RudderStack dashboard.

Databricks

Concurrency does not work for cleanup.

Other issues

Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
pb insert does not work for Redshift, Databricks, and BigQuery.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
Timegrains is an experimental feature. There might be some undiscovered issues.

Version 0.17

14 August 2024
Schema version: 72

What’s New

pb show plan: A new subcommand is added to show detailed information about the materials with different timegrains along with their dependencies in the order of execution.
The casing for the input tables names defined using double quotes is preserved in the warehouse. For example, the table defined as table: "TableName" will be referred as TableName in the warehouse.
Schema has been updated from 71 to 72.

Improvements

Relevant error message is shown if the computed status for a material is not possible.
An error is thrown if multiple feature views are defined with the same name.

Bug Fixes

Fixed the issue where input_vars was not getting defined on the packages.
Resolved the bug where the project was failing due to empty or commented YAML files.

Known Issues

BigQuery

pb validate access command does not work for BigQuery.

Redshift

If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
Cross database references can fail on Redshift for a few clusters.
While creating Activations, validation for Redshift does not work correctly in the RudderStack dashboard.

Other issues

Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
pb insert does not work for Redshift, Databricks, and BigQuery.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
The code for validity_time is redundant and should be removed.
Timegrains is an experimental feature. There might be some undiscovered issues.

Version 0.16.1

6 August 2024

Bug Fixes

Fixed the concurrency issue where project run was failing because the same ID stitcher model was served as an input source to two models.
Resolved the ambiguous column issue by using an alias for selecting main_id in the entity_var.

Version 0.16

1 August 2024
Schema version: 71

What’s New

You can now choose between the username-password and key-pair authentication modes while running the pb init connection command for Snowflake.
Added support to remove materials based on time in hours and milliseconds. For example, to remove materials older than 3 hours, use pb cleanup materials --retention_time_in_hours 3. To remove materials older than 100 milliseconds, use pb cleanup materials --retention_time_in_ms 100.
Schema has been updated from 69 to 71.
Edge type preferred has been renamed to coercive. Here’s a sample code:

 - name: example
   model_type: sql_template
   model_spec:
     single_sql: |
       select * from {{this.DeRef(test, dependency = "coercive")}}

Improvements

Changed the casting for timestamp columns for Snowflake, Redshift, and Databricks warehouses to avoid a timestamp with string comparison. CAST(timestamp_column AS TIMESTAMP) < ‘2024-07-28T23:00:00Z’ has been changed to CAST(timestamp_column AS TIMESTAMP) < CAST(‘2024-07-28T23:00:00Z’ AS TIMESTAMP)

Bug Fixes

Fixed the Sequence number conflict found error on BigQuery.
Fixed the bug in RudderStack dashboard where entities weren’t showing up, in case no features were defined in the project.
Resolved the intermittent timeout waiting for python client to initialize error for projects using pynative models.
Fixed the directory not empty error when Profiles tries to clone a git repository.
Fixed the issue where the registry entry for materials was not using the same casing due to which the relations were getting created in the warehouse. Now the cleanup is able to delete materials that were skipped earlier due to case mismatch.
Fixed the issue where the same hash was calculated for different cohorts, as the filter pipeline wasn’t being considered.
Fixed the issue where some procedures on Snowflake and BigQuery were not getting removed after the run.

Known Issues

RudderStack does not support accessing input sources in a different project for the BigQuery warehouse.
Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
pb validate access command does not work for BigQuery.
pb insert does not work for Redshift, Databricks, and BigQuery.
Cross database references can fail on Redshift for a few clusters.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
The code for validity_time is redundant and should be removed.
Timegrains is an experimental feature. There might be some undiscovered issues.
While creating Activations, validation for Redshift does not work correctly in the RudderStack dashboard.

Version 0.15

10 July 2024
Schema version: 69

What’s New

Support for key-pair authentication in Snowflake for enhanced security.
Schema has been updated from 67 to 69.

Improvements

Replaced ROW_NUMBER() OVER(ORDER BY 1) with ROW_NUMBER() OVER() for Redshift and BigQuery. This eliminates an unnecessary ORDER BY first column.

Known Issues

When using the CLI: You will have to manually make changes in the siteconfig.yaml file to add a Snowflake connection with key-pair authentication.
RudderStack does not support accessing input sources in a different project for the BigQuery warehouse.
Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
pb validate access command does not work for BigQuery.
pb insert does not work for Redshift, Databricks, and BigQuery.
Cross database references can fail on Redshift for a few clusters.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
The code for validity_time is redundant and should be removed.
Timegrains is an experimental feature. There might be some undiscovered issues.
While creating Activations, validation for Redshift does not work correctly in the RudderStack dashboard.

Version 0.14.2

28 June 2024

Bug Fixes

Resolved “authentication token has expired” issue on Snowflake.

Version 0.14.1

26 June 2024

Bug Fixes

Resolved few migration issues.
Fixed BigQuery issues where the project name had special characters.

Version 0.14

19 June 2024
Schema version: 67

What’s New

You can now refer to entity_vars defined on cohorts using the <cohort_name>.Var("<var_name>") syntax along with the original entity.Var("v1") syntax.
Vars from ancestor vars are now inherited in derived cohorts. If a derived cohort has vars with the same name as ancestors, they are overridden.
You can also specify timegrains on a specific entity_var and not only on the entire var_group.
If a column is specified as mandatory in a model contract, it must be present in the warehouse. Otherwise, the project will fail. Also, contract validation is now enabled for SQL models and Python models.
pb show models command now displays additional information such as the enable_status of models, warehouse name, namespace, and timegrains.
In materialization status, the flag not_needed has been renamed to only_if_necessary.
Some experimental features are added to optimize performance by partitioning var tables.
A new key row_identifier (experimental) has been added under app_defaults in the inputs.yaml file.
Schema has been updated from 63 -> 67 in the pb_project.yaml file.

Improvements

Few optimizations in the RudderStack dashboard to make it leaner and faster.

Bug Fixes

Fixed the issue where features list wasn’t getting added to input models build specification.
Resolved a bug where <entity>/feature_views model wasn’t getting created, in case the feature_views key didn’t have any value.
Resolved the error where some BigQuery projects were failing with the message Dataset <> not found in location.
Fixed the symlink issue where the pb run/pb compile commands were failing when the relative path of the project is passed.
Resolved the bug where input models were getting enabled if they did not exist earlier, even if their status wasn’t set.
Error is now being thrown in case all the inputs of a model are disabled but they are mandatory.
Fixed the error nil pointer dereference when other_id is not defined in the feature views.
Resolved the issue where different cohorts with same feature names were throwing an error.
Fixed the bug where a derived cohort was not able to refer other cohort’s vars.
Fixed column name is ambiguous error in feature views when there are vars with the same name in different timegrains.
Resolved the issue where some projects with concurrency were failing with error concurrent map read and map write.
Fixed the issue with force runs on models with begin time.

Known Issues

RudderStack does not support accessing input sources in a different project for the BigQuery warehouse.
Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
pb validate access command does not work for BigQuery.
pb insert does not work for Redshift, Databricks, and BigQuery.
Cross database references can fail on Redshift for a few clusters.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
The code for validity_time is redundant and should be removed.
In some cases, you may need to install the profiles-rudderstack and profiles-rudderstack-bin pip packages separately.
You may have to execute the compile command once before executing validate access. Otherwise, you will get a seq_no error.
Timegrains is an experimental feature. There might be some undiscovered issues.
Activations does not work with Redshift warehouse.

Version 0.13.2

11 June 2024

Bug Fixes

Fixed issues with auto-migrate on Windows.
Resolved insufficient memory error encountered while adding main_id to input_var tables in Redshift.

Version 0.13.1

23 May 2024

Bug Fixes

Resolved the issue with remapped models in entity projects in the RudderStack dashboard.
Fixed timestamp casting issues on BigQuery.

Known Issues

RudderStack does not support accessing input sources in a different project for the BigQuery warehouse.
Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack gives an error during cleanup when trying to drop views created by the other user, like user_var_table.
pb validate access command does not work for BigQuery.
pb insert does not work for Redshift, Databricks, and BigQuery.
Cross database references can fail on Redshift for a few clusters.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually remove the entire folder from WhtGitCache to make it work.
The code for validity_time is redundant and should be removed.
In some cases, you may need to install the profiles-rudderstack and profiles-rudderstack-bin pip packages separately.
You may have to execute the compile command once before executing validate access. Otherwise, you will get a seq_no error.
Cohort features do not inherit the parent’s features.
RudderStack dashboard lists a cohort only if it has features.
Timegrains is an experimental feature. There might be some undiscovered issues.
Activations does not work with Redshift warehouse.

Version 0.13

16 May 2024
Schema version: 63

What’s New

Timegrains feature (experimental) lets you reuse the output material of a model within the specified time period, preventing unnecessary recalculations⁠. You can define its value as a day, week, month etc. to compute features at the end of that particular period.
Cohorts feature (experimental) lets you define core customer segments within an entity based on some characteristics.
Removed support for rebase_incremental in project_spec. However, you can specify it using the command-line tool.
The output folder now generates artifacts for run and compile commands in separate subfolders with the same name.
pb init pb-project now comes with a sample SQL model.
ReadFile support introduced in YAML which lets you highlight syntax. In a SQL model, you can use {{this.ReadFile("models/sql_file.sql")}} to refer SQL content from a file.
Material registry is now at v5. It includes two new columns - model_ref and registry_version. The registry will be automatically migrated once you execute pb run command.

Improvements

Improved performance in compilation and ID stitching processes, resulting in faster operations.
Improved error handling at some places.
Few internal refactorings for enhanced working.

Bug Fixes

Removed include_untimed key from the pb_project.yaml file as it was redundant.
Fixed sporadic lengthening of project runs in RudderStack dashboard.

Known Issues

Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack will throw error during cleanup when trying to drop views created by the other user, such as user_var_table.
pb validate access command does not work for BigQuery.
Some commands such as pb insert does not work for Redshift, Databricks, and BigQuery.
For a few clusters, cross database references can fail on Redshift.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually clear WhtGitCache folder to make it work.
The code for validity_time is redundant and should be removed.
In some cases, you may need to install the profiles-rudderstack and profiles-rudderstack-bin pip packages separately.
You may have to execute the compile command once before executing validate access. Otherwise, you will get a seq_no error.
Cohort features do not inherit parent’s features.
RudderStack dashboard lists a cohort only if it has features.
Timegrains is an experimental feature. There might be some undiscovered issues.
Activations does not work with Redshift warehouse.

Version 0.12.1

2 May 2024

Improvements

If all the inputs of a model are disabled, the model is disabled by default.

Bug Fixes

Resolved bug where some projects were failing in case nested columns were missing.
Updated the migration logic to preserve old view names. The reason being that some existing Activation API projects were failing due to the renaming of serve_traits to feature_views.

Known Issues

Linux users might see this warning for all command runs - you can ignore it: WARN[0000]log.go:228 gosnowflake.(*defaultLogger).Warn DBUS_SESSION_BUS_ADDRESS envvar looks to be not set, this can lead to runaway dbus-daemon processes. To avoid this, set envvar DBUS_SESSION_BUS_ADDRESS=$XDG_RUNTIME_DIR/bus (if it exists) or DBUS_SESSION_BUS_ADDRESS=/dev/null.
Redshift: If two different users create material objects on the same schema, RudderStack will throw error during cleanup when trying to drop views created by the other user, such as user_var_table.
pb validate access command does not work for BigQuery.
Some commands such as pb insert does not work for Redshift, Databricks, and BigQuery.
For a few clusters, cross database references can fail on Redshift.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually clear WhtGitCache folder to make it work.
The code for validity_time is redundant and should be removed.
In some cases, you may need to install the profiles-rudderstack and profiles-rudderstack-bin pip packages separately.
You may have to execute the compile command once before executing validate access. Otherwise, you will get a seq_no error.

Version 0.12

25 April 2024
Schema version: 61

What’s New

Support for external tables on Redshift Serverless.
Redshift users can now also connect via SSH Tunnel.
Support to specify the version on which you want to run your Profiles project. It is also backported to v0.10.8 and v0.11.5 onwards.
Feature views path refs look like below:
- user/all/feature_view # By default, the id served is user main_id.
- user/all/feature_view/using_email # For non default IDs, the path ref has using_<idname>
- user/all/feature_view/salesforce_id_stitched_features # The names can still be overridden.

Ability to add description (optional) to feature tags in the project file.

available_tags:
    - name: demographics
      description: all tags related to user demographics
    - name: billing

Reorganized the overall flow when defining a Redshift connection. You can specify the region while using Redshift.
Creating a new project using pb init pb-project now comes with sample data in CSV format, in a folder named csvs. These are referenced in the input file.
Renamed Entity Traits 360 model to Feature Views. Also, in the pb_project.yaml file, a Feature Views model is included by default under the entities definition.
Schema has been migrated from 54 -> 61 in the project file.

Improvements

pb init connection now fetches connection name from the site configuration file, and asks you to specify one as well. If you don’t enter any value and there’s only one entry in the file, then it picks that value.
Default ID stitcher now uses all package models for ID sources as well, in addition to local ones.
Relevant error is now thrown, if ID types for an entity has not been specified in the pb_project.yaml file.
An entity project in the RudderStack dashboard no longer fails in case a few columns are missing in the source tables.
Debugging is improved and the correct line number is displayed.
pb show models command now includes model dependencies along with some more statistics.
Few retoolings from the ground up.

Bug Fixes

Resolved bug related to failure when a git repo couldn’t be found in the RudderStack dashboard.
Resolved bug where default ID stitcher was getting created even if there were no ID edges.
Resolved bug where Redshift Profiles projects were failing if names of entity_var/input_var were longer than 47 characters.
Resolved bug where pb run --rebase_incremental command was taking edges from previous runs.
Resolved the issue where a project wasn’t compiling in case it referenced a package having inputs that weren’t defined in the application with the same name.

Version 0.11.5

16 April 2024

Bug Fixes

Fixed issue for Redshift where the driver version wasn’t getting populated correctly.
Improved cleanup functionality for Redshift by dropping procedures that were used for creating entity_vars.
Resolved the issue where migrated project folder’s files weren’t getting deleted.
Fixed the bug where pb run --rebase_incremental command was taking edges from previous runs.
Few internal refactorings while returning data types of columns.

Version 0.11.3

1 April 2024

What’s New

An optional parameter column_data_type to specify the data type for an entity_var/input_var.
Support for programmatic credentials for Redshift.
Schema update in the project yaml file from 53 to 54.

Improvements

Better error propagation in case of concurrency.
Few internal refactorings for improved overall working.

Bug Fixes

Resolved relation still open error when accessing external tables in Redshift.
Fixed some bugs when getting the latest seq_no for a material in BigQuery.
Resolved the issue of conflict in row-ID in case of very large datasets in BigQuery.
Begin and end time of all models are now in UTC timezone. This fixes a few inconsistency issues in models.
Resolved a concurrency issue which occurred on two different root models with the same name.

Version 0.11.2

15 March 2024

What’s New

You can now do parallel processing while running a project using the --concurrency flag. Currently, this is supported only for Snowflake warehouse. It is recommended to use this option judiciously as applying a very large value can impact your system resources.
RedShift users can now access external tables in their data catalog(s).

Improvements

Project created using pb init pb-project now works for all warehouses.

Bug Fixes

Fixed issues encountered while running BigQuery projects on Windows.
Resolved errors for entity var names in case they match with input column name.
Resolved bugs related to inserting seq_no.

Version 0.11.1

7 March 2024

Includes bug fixes related to creating vars on ID models and nil model remapping.

Version 0.11

1 March 2024

What’s New

RudderStack now supports BigQuery (beta), offering the same seamless experience as on other data warehouses.
CSV models (Experimental): In the inputs specs, RudderStack has added the ability to read data from a CSV file, instead of a Database table/view. You can use files from local storage, or kept on S3. Under app_defaults, instead of table/view, use csv (local storage) or s3 (kept on S3) followed by the path where the CSV file is kept. Note that this feature is experimental, and RudderStack currently supports S3 on Snowflake and Redshift. A sample code is as follows:

    app_defaults:
      csv: "../common.xtra/Temp_tbl_a.csv"
…
    app_defaults:
      s3: "s3://s3-wht-input-test-bucket/test/Temp_tbl_d.csv"

Filter IDs: You can now filter out a vast number of ID’s using SQL. For example, if you wish to exclude all blacklisted ID’s that are listed in an input model named csv_email_blacklist and user ID’s from an SQL model named sql_exclusion_model, then, you may edit your project file as:

id_types:
  - name: email
    filters:
      - type: exclude
        sql:
          select: email
          from: inputs/csv_email_blacklist
  - name: user_id
    filters:
      - type: exclude
        sql:
          select: user_id
          from: models/sql_exclusion_model

Pre and Post Hooks: A pre hook enables you to execute an SQL, before running a model, for example, if you want to change DB access, create a DB object, etc. Likewise, a post hook enables you to execute an SQL after running a model. The SQL can also be templatized. Here’s an example code snippet:

models:
  - name: test_id_stitcher
    model_type: id_stitcher
    hooks:
      pre_run: "CREATE OR REPLACE VIEW {{warehouse.ObjRef('V1')}} AS (SELECT * from {{warehouse.ObjRef('Temp_tbl_a')}});"
      post_run: 'CREATE OR REPLACE VIEW {{warehouse.ObjRef("V2")}} AS (SELECT * from {{warehouse.ObjRef("Temp_tbl_a")}});'
    model_spec:

pb show models - You can now view in JSON format by passing the flag –json.
For Databricks, RudderStack now supports the pb validate access command.
RudderStack has reverted to having a custom ID stitcher in a new project created using pb init pb-project.
When creating a new connection in Redshift, you’ll now be asked to input sslmode. You can enter either disable (default) or require. This will help RudderStack’s tool to work with Redshift DB’s that require SSL mode to be enabled.
RudderStack supports triggering tasks by using URL and also read the status back.
RudderStack dashboard now supports Git Projects hosted on BitBucket and GitLab.
In model specs, a materialization’s enable_status is changed to snake_case. That is, enable_status: mustHave -> enable_status: must_have.
Schema version in the project file has been updated from 49 to 53.

Improvements

Better error messages are shown in case of incorrect/missing/duplicate entity-vars.
Error handling has been improved at a few places in Python models.
Model path refs are now case insensitive.
The command pb migrate auto can now handle the case where model folders aren’t present.
Specific messages are now shown, in case of errors in the material registry.
Due to limitations of Databricks SQL, RudderStack has added restrictions on using catalog hive_metastore. So, in case a user on that catalog tries to use RudderStack’s tool, an error is thrown.

Bug Fixes

Resolved the intermittent issue in Redshift where it throws an error ptr_to_latest_seqno_cache does not exist.
Bugs in pb show idstitcher-report, pb show user-lookup, and pb cleanup materials commands have been rectified. pb show idstitcher-report is still flaky, however, RudderStack team is working on improving it.
Fixed bug in packages wherein entityvars/inputvars weren’t able to refer SQL models.
Resolved erroneous queries for validate access command in case of missing privileges.
Recsolved the issue where git repo wasn’t getting cloned in case cache_dir in siteconfig was written using tilde notation.
Fixed some bugs related to begin_time of models.
Resolved a few issues when cloning Git projects in the web app.
Several fixes in gRPC, making it more stable.
The remapping: key is removed (if exists) in models/inputs.yaml as it was redundant.
Resolved some bugs in incremental ID stitcher.

Known Issues

pb validate access command does not work for BigQuery.

Version 0.10.6

19 January 2024

An internal fix to address issues that arose from a recent update by Snowflake.

Version 0.10.5

9 January 2024

Some internal fixes to make py-native models more robust.

Version 0.10.4

15 December 2023

Our latest version has a plethora of features that makes our product more feature-rich and impactful.

What’s New

Vars as models: Earlier, Vars could only be defined inside the feature table under vars: section. Now, Vars are defined independent of feature tables. In the model specs file, we have created a new top level key called var_groups. We can create multiple groups of vars that can then be used in various models (eg. in feature table). All vars in a var-group need to have the same entity. So if you have 2 entities, you need at least 2 var groups. However, you can create multiple var_groups for every entity. For example, you can create churn_vars, revenue_vars, engagement_vars etc. So that it is easier to navigate and maintain the vars that you need. Each such model shall have name, entity_key and vars (list of objects). This is in line with Profiles design philosophy to see everything as a model.
User defined model types via Python [Experimental feature]: Ever wondered what it would take to implement a new model type yourself? Custom model types can now be implemented in Python. Check out this library for some officially supported model types with their Python source. Note that this is an experimental feature, so the implementation and interfaces can change significantly in the upcoming versions. To use a python package model in your project, simply specify it as a python_requirement in pb_project.yaml, similar to requirements.txt. The BuildSpec structure is defined using JSON schema within the Python package. Below code snippet shows how the requirements such as for training and config can be specified in the project:

    entities:
      - name: user
      python_requirements:
      - profiles_rudderstack_pysql==0.2.0 #registers py_sql_model model type

    models:
      - name: test_py_native_model
        model_type: py_sql_model
        model_spec:
          occurred_at_col: insert_ts
          train_config:
            prop1: "prop1"
            prop2: "prop2"

Default ID stitcher: Until now, when a new project was created using pb init pb-project, the file profiles.yaml had specifications for creating a custom ID stitcher. That has a few limitations, when edge sources are spanning across packages. Also, we observed that several of our users weren’t doing much changes to the ID stitcher, except for making it incremental. As a solution, we have a “default ID stitcher”, that is created by default for all projects. It runs on all the input sources and ID types defined. For quickstart purposes, users needn’t make any changes to the project, to get the ID stitcher working. In case any changes are to be made, then a user can create a custom ID stitcher, as was done in earlier versions.
Default ID types: Now, common concepts like ID types can be loaded from packages. So we needn’t define them in all new projects. Hence, we have moved the common ID type definitions into a library project called profiles-corelib. So when you create a new project, the key id_types is not created by default. In case you wish to create a custom list of ID types that is different from the default one, then you may do it as was the case in earlier versions.
Override packages: Continuing from previous point: packages now have overrides materialization spec. In case you wish to add custom ID types to the default list or modify an existing one, then you may extend the package to include your specifications. For the corresponding id_type, add the key extends: followed by name of the same/different id_type that you wish to extend, and corresponding filters with include/exclude values. Below is an example of the same:

    packages:
        - name: foo-bar
        url: "https://github.com/rudderlabs/package-555"
    id_types:
        - name: user_id
        extends: user_id
        filters:
            - type: exclude
              value: 123456
    id_types:
        - name: customer_id
        extends: user_id
        filters:
            - type: include
              regex: sample

entity_var tags: You can now define a list of tags in the project file under tags: key. Then, you can add a tag to each entity_var.
Redshift: We have added support for the RA3 node type. So now our users on that cluster can cross-reference objects in another database/schema.
Schema version in the project file has been updated from 44 -> 49.

Improvements

Generated ID’s are now more stable. This means that they are unlikely to adapt to merging of ID Clusters, thereby creating a more accurate profile of your users.
By default, every entity_var is a feature, unless specified otherwise using is_feature: false. So now, you need not explicitly add them to the features: list.
You can now add escape characters to an entity_var’s description.
Several internal refactorings to improve overall working of the application.

Bug Fixes

An entity_var having a description with special characters was failing during project re-runs. This has now been resolved.
We have fixed the bug where two entity_vars across different entities in the same project couldn’t have the same name.
Fixed some bugs related to vars as models, auto migration of projects, and ID lookup.

Known Issues

Redshift: If two different users create material objects on the same schema, then our tool will throw error when trying to drop views created by the other user, such as user_var_table.
Some commands such as insert do not work on Redshift and Databricks.
For a few clusters, cross DB references can fail on Redshift.
If you are referring a public package in the project and get ssh: handshake failed error, then you’ll have to manually clear WhtGitCache folder to make it work.
The code for validity_time is redundant and should be removed.
Sometimes you may have sometimes install both the pip packages separately (profiles-rudderstack and profiles-rudderstack-bin).
You may have to execute the compile command once, before executing validate access. Otherwise, you can get a seq_no error.

Version 0.9.4

8 November 2023

This release includes the following bug fixes and improvements:

pb run --grep_var_dependencies - we are now setting default values using the rule “if a project is migrated on load from a version older than 43, then grep_var_dependencies will default to true otherwise false”. Also, handled a null pointer case for non existent vars listed in dependencies.
pb migrate_on_load / migrate auto - we have made the message clearer on curly braches in dot syntax message.
pb migrate manual - we have removed compatibility-mode as it was no longer required.
A few internal refactorings.

Version 0.9.3

2 November 2023

This release addresses a few vulnerability fixes.

Version 0.9.2

26 October 2023

In case you are unable to install then we recommend having Python3 versions from 3.8 to 3.10.

This release includes a bug fix on self dependency of vars, in case column has same name as entity-var.

Version 0.9.1

19 October 2023

Our latest release contains some useful features and improvements, as asked by our users.

After the auto-migration to v44, you might be shown some warnings to do changes in the YAML. Please check the Tutorials section. Or, you may contact our team and we will assist you with the same.

What’s New

We have added support for Databricks (beta). Now Databricks users can seamlessly create ID stitcher and feature table models, without writing complex SQL queries! If you’re using Databricks and want to try out Profiles, kindly get in touch with our team.
Vars as models : Now, entity_vars and input_vars can be treated as independent models. Presently, they are tied to a feature table model. In SQL template text, for example in SQL model templates, please use {{entity-name.Var(var-name)}} going forward to refer to an entity-var or an input-var. For example, for entity_var user_lifespan in HelloPbProject, change select: last_seen - first_seen to select: '{{user.Var("last_seen")}} - {{user.Var("first_seen")}}'.
pb show dataflow and pb show dependencies commands - A new flag --include_disabled flag is added to let disabled models be part of the generated image. Also, we now show the relative path from local root, instead of the full path.
pb run command - Added flag --ignore-model-errors to let the project continue running in case of an erroneous model. So, the execution wouldn’t stop due to 1 bad model.
pb run - Added flag --grep_var_dependencies (default: true) which searches for vars dependencies by using grep over fields from vars definition.
pb show idstitcher-report - Added flag --seq_no, using which a specific run for an ID stitcher model can be specified.
Best schema version - For a library project, in the url key of packages, we have introduced the concept of “best version tag”. That is, instead of specifying the specific Git URL of the library project, we give a URL with GIT tag url: https://github.com/rudderlabs/librs360-shopify-features/tag/schema_{{best_schema_version}}. Using this will make our tool use the best compatible version of the library project, in case of any schema updates.
Schema has been migrated from version 42 -> 44.

Improvements

The command pb show user-lookup now includes more details including the count of rows created and total number of features.
Commenting out features will ensure that the corresponding entity-var and any related entity-var/input-var being used only for computation of this commented feature wont run
Several improvements done beneath the surface.

Bug Fixes

The flag -- force was having issues in dropping priorly created materialization models. This has now been resolved.
Fixed bug where project was unable to run due to giving a custom name to the ID stitcher.
Resolved an issue in the command pb show idstitcher-report, in the case if the hash of the ID Stitcher model has changed from that of the last run, rerunning the ID Stitcher model.
Removed flag -l from the command pb show idstitcher-report as it was redundant.

Known Issues

Redshift: If two different users create material objects on the same schema, then our tool will throw error when trying to drop views created by the other user, such as user_var_table.
Some commands such as insert do not work on Redshift and Databricks.
For a few clusters, cross DB references can fail on Redshift.

Version 0.8

25 August 2023

What’s New

Model Contracts - We have added support for model contracts and their validation. For every input or SQL model, there’s a new key contract: which contains the following keys: is_optional (boolean, to indicate if the model is optional), is_event_stream (boolean, in case the data is event stream and has timestamp), with_entity_ids (list of all entities model contains), with_columns (list of all column names model have). A contract can be passed along with the model path in this.DeRef. For more information, check out Model Contracts.
Inputs model - The keys occurred_at_col and ids are now a part of app_defaults, to reinforce that they can also be overridden.
Schema has been migrated from 40 -> 42 in the project file.

Improvements

The command pb cleanup materials now removes tables generated by Python models also.
pb show user-lookup now includes user traits from Python models as well.
A few changes under the hood, for more efficient processing and execution.

Bug Fixes

Fixed issue in Python models where validity of the train file wasn’t working and it so was retraining the model(s) on every run.
Resolved the bug where wrong credentials in siteconfig file was not printing the exact error.
Queries for checking warehouse access (grant) were duplicated and therefore recursively checking grants on the same models again and again. This resulted in taking more time than what was required. It has now been fixed.
pb migrate auto - There was an issue in migration of multi-line strings of SQL models, that has now been resolved.

Version 0.7.3

14 August 2023

What’s New

pb show idstitcher-report:pb show idstitcher-report: By passing flag --id_stitcher_model, you can now create an HTML report with relevant results and graphics including largest cluster, ID graph, etc.
Material Registry has been updated to version 4, as additional information is now stored for target (as defined in siteconfig), system username, and invocation metadata (hostname and the project’s invocation folder). So now, if anyone logs into the system and creates material objects using PB, then these details will be stored. This is based on a feature request from one of our customers. Note: make sure to execute pb validate access for migrating the registry.
pb discover materials:pb discover materials - This command now shows a few additional columns - target, username, hostname, invocation folder.
Default ID stitcher: In the inputs file, the key to_default_stitcher needs to be set to true explicitly for an ID to get picked in the default ID stitcher. This field is optional and by default set to false, without impacting if the project is using a custom ID stitcher. In your project file, if you remove the key id_stitcher: models/<name of ID stitcher model>, then it’ll use the default ID stitcher and create a material view of the name <entity_name>_default_id_stitcher.
In the inputs.yaml file, table or view names now appear under a key named app_defaults:. This signifies that these are values that input defaults to, when the project is run directly. For library projects, inputs can be remapped and appdefaults overridden. when library projects are imported.
Schema has been migrated from 38 -> 40 in the project file.

Improvements

pb init pb-project:pb init pb-project: Added keys on default ID stitcher.
A few improvements behind the scenes, for enhancing the overall functionality.

Bug Fixes

Resolved the issue where projects migrated using migrate_on_load were referring to the location of the migrated project in the material registry. This was affecting the count of ID’s before and after stitching.
Fixed bug where ID stitcher wouldn’t check whether a material was actually existing in the database, before running in incremental mode.
When the material registry was on an unsupported common tables version, then the project environment loading would fail, thereby crashing the application. This has now been resolved.
Features defined in Python models, now do appear in the list of features.
Vars can still be specified in specs of a feature table model. However, the app ignores them. This is a bug and would be fixed in subsequent releases.

Version 0.7.2

24 July 2023

Our newest release brings enhanced functionality and a more efficient experience.

What’s New

Model Enable/Disable: You can now enable or disable specific models using the materialization key in model specifications. Use the status key to set values. For more information, refer to Models enabling themselves.
Migrate Auto: When migrating a project, the ordering of elements now remains the same as in the original files, preserving comments.
Graceful Application Exit: You can now exit the application gracefully while it’s running. For example, if you’re generating material tables using the run command, you can exit using Ctrl+C.
Schema Migration: The schema version in the project file has been updated from 37 to 38.

Improvements

Projects created using init pb-project now include dependencies.
Instead of generating one big SQL file, we now create multiple files in a folder during SQL generation of a feature table model. This reduces the disk space requirements.
Internal optimizations have been implemented to improve overall performance and efficiency.

Bug Fixes

An issue has been fixed where insufficient grants for accessing the warehouse would lead to duplicate suggested queries. Also, in some cases, incorrect queries were displayed, such as when a Redshift user was asked to grant a role.
The project URL is now being stored in the material registry, instead of GitHub passkey.
Fixed a bug where macros defined in a separate file as global macros were unable to access a common context.
Resolved a bug where Python models were not appearing in the dependency graph.

Version 0.7.1

23 June 2023

Our latest release addresses some critical issues in the previous release. Therefore, if you’re on v0.7, then it’s highly recommended to update to the latest version.

Version 0.7

22 June 2023

Our newest release is quite significant in terms of new features and improvements offered. Be sure to try it out and share your feedback with us.

What’s New

query - A new command which displays output of tables/views from the warehouse. So you can view generated material tables from the CLI itself. For example, pb query "select * from {{this.DeRef("models/user_id_stitcher")}}".
show idstitcher-report - A new sub command that creates report on an ID stitcher run. Such as, whether it converged, count of Pre-stitched ID’s before run, Post-stitched ID’s after run, etc. Usage: pb show idstitcher-report .
show user-lookup - A new sub command that allows you to search a user by using any of the traits as ID types. E.g., pb show user-lookup -v <trait value>.
If non-mandatory inputs required by the model are not present in the warehouse, you can still run the model. Applicable to packages and feature tables.
Schema updated from 33 -> 37 in the project file. Please note that the material registry has been migrated to version 3, so you’ll have to execute pb validate access once in order to execute the run command.

Improvements

Added an optional field source_metadata in the model file inputs.yaml.
Added EnableStatus field in materialization so that models can be enabled and disabled automatically based on whether it is required or not.
Default ID stitcher now supports incremental mode as well.
In macros, you can now specify timestamps in any format.

Bug Fixes

In case a project is migrated using flag migrate_on_load, then src_url in the material registry was pointing to the new folder. Now, that is fixed.
Resolved bugs in generating edges for dependency graphs.
Tons of several other improvements and bug fixes under the hood.

Version 0.6

26 May 2023

We are excited to announce the release of PB Version 0.6: packed with new features, improvements, and enhanced user experience.

What’s New

Support for begin time and end time: Our latest release introduces the ability to specify time range for your operations, using two new flags --begin_time and --end_time. By default, the --end_time flag is set to now. For example, you can use the command pb run --begin_time 2023-05-01T12:00:00Z to fetch all data loaded after 1st May 2023. Note that the flag --timestamp is now deprecated.
A new flag, model_refs, has been introduced which restricts the operation to a specified model. You can specify model references, such as pb run --model_refs models/user_id_stitcher.
seq_no - Another new flag, using which you can Continue a previous run by specifying its sequence number. Models that already exist would not be rebuilt, unless --force is also specified.
Show command - pb show dependencies has been added to generate a graph showcasing model dependencies. This visual representation will help you understand the relationships between various models in your project.
Show command - pb show dataflow: Another new command which generates a graph with reversed edges, illustrating the dataflow within your project.
migrate_on_load - A new flag, migrate_on_load, has been introduced. When executing the command pb run --migrate_on_load, by default this flag creates a migrations folder inside the project folder that has migrated version of your project to the latest schema version and runs it on the warehouse. This simplifies the process of upgrading your project to the latest schema version, without changing source files.
migrated_folder_path - Continuing from previous command, you can use this flag to change folder location of the migrated project.
Schema in the project file has been updated to version 33.

Improvements

SQL Models now provide more relevant and informative error messages instead of generic “not found” errors. This simplifies troubleshooting and debugging processes.
Numerous improvements and optimizations have been made behind the scenes, enhancing the overall performance and stability of PB.

Version 0.5.2

5 May 2023

Our latest release offers significant performance improvements, enhancements, and bug fixes to provide a better experience.

What’s New:

A new command, pb show models, which displays various models and their specifications in the project.
Ability to exit the application while the run command is being executed.
Project schema version has been migrated to 30.

Improvements:

Major performance improvements for Redshift. In large data sets, it will reduce the time taken to create ID stitcher table to less than 1/4th of the time taken earlier.
insert command now picks the connection specified in the current project folder. If not available, it picks “test” in the connection file.
Siteconfig is now validated when project is loaded.
The cleanup materials command now removes SQL models as well.

Bug Fixes:

Resolved the problem where values with null timestamps were excluded from incremental ID stitcher.
The insert command was showing a success message even if no tables were inserted in the warehouse. This has been fixed.

Version 0.5.1

11 April 2023

What’s New

Updated schema to version 28 in the project file.

Improvements

Changed project path parameter from -w to -p for improved usability.

Bug Fixes

Addressed a few reported bugs for an improved user experience.
Implemented performance enhancements to optimize overall system performance.

Version 0.5

28 March 2023

This release offers significant new additions and improvements, as well as bug fixes.

What’s New

Cleanup materials - You can now use the command pb cleanup materials to delete materials in the warehouse automatically, without the need for manual deletion. Just specify the retention time period in the number of days (default 180) and all tables/views created prior to that date will be deleted.
Schema has been migrated to 27. This includes the following changes:
- pb_project.yaml - The schema version has been updated from 25 → 27. Also, main_id is removed from id_types as main_id_type is now optional, rudder_id is the main_id_type by default.
- models/profiles.yaml - To explicitly declare edge source ids, each value in edge_sources now requires a from: key to be appended. Also, if you didn’t define main_id in the project file, then no need to specify here.

Improvements

In the backend code we’ve enabled registry migration which flattens the registry, enabling incremental ID stitcher to operate on incomplete materials. It also introduces a mechanism for migrating common tables.
We have implemented better error handling for cases where an incorrect model name is passed. Any errors related to incorrect model names are now properly identified and handled by the system.
Based on feedback from our users, we have renamed default models from domain_profile<> to user_profile<>.

Due to changes in registry, we will be depricating older versions of PB.

Bug Fixes

Fixed the bug where some experimental features, such as Discover, were not working for Redshift.
Addressed the problem where validation errors were incorrectly being triggered when a connection had multiple targets, one of which was invalid. The system now only generates an error if the warehouse target that is being passed has errors.
In addition to previous one, a few more bugs were fixed that were related to validation.
Errors were coming for users who had initialized the GIT repository but had not added the remote origin. This issue has now been fixed.

Known Issues

Warning: While the run command is being executed, canceling it by pressing Ctrl+C doesn’t work as expected. Though it will stop the program’s execution on the CLI, the query will keep running on the data warehouse. This is a documented Snowflake behavior.
In a model, an input can’t use columns named “MAIN_ID”, “OTHER_ID”, “OTHER_ID_TYPE”, or “VALID_AT” in its ID SQL.
When creating a connection via init command, pressing the Ctrl+C command doesn’t exit the application.
migrate auto jumbles up the order and removes comments.
On Redshift, validate access passes all tests, but run command sometimes fail giving error “permission denied for language plpythonu”.
Some commands such as insert do not work on Redshift.
For a few clusters, cross DB references can fail on Redshift.
The command migrate auto migrates siteconfig in your home directory but not any local one.
While working with same type of data in Snowflake and Redshift you might encounter errors where it works on Snowflake but not on Redshift. This is due to the fact that implicit casting of different data types for different function or operator might not be supported on one data warehouse while supported on other.

Version 0.4

2 March 2023

We are proud to announce the latest version of PB 0.4, which includes several new features and improvements.

What’s New

Redshift - We are excited to share that we now offer Redshift integration. With YAML, you can now effortlessly create ID stitched and feature table models on your Redshift warehouse, without any difficulty.
Incremental ID stitching: You can now stitch together data from multiple sources in incremental mode. When new data is added to your source tables, only that data will be fetched, without needing to reload the whole table each time! This shall result in significant performance improvements from earlier versions, especially if the delta of new data is much smaller compared to what’s already been stitched.
Insert: A new command, allowing users to add sample data to their warehouse, without having to manually add them.
Schema has been migrated to 25. This includes the following changes:
- models/profiles.yaml - Renamed entityvar to entity_var and inputvar to input_var
- pb_project.yaml - Renamed profile to connection.
- siteconfig.yaml - Renamed profiles to connections.

Be sure to use the migrate auto command to upgrade your project and the connections file.

Improvements

The command init profile has been renamed to init connection.
Lots of modifications under the hood.

Bug Fixes

Resolved issue on default values in an entity var, ensuring that the values are properly set.

Known Issues

Warning: While the run command is being executed, canceling it by pressing Ctrl+C doesn’t work as expected. Though it will stop the program’s execution on the CLI, the query will keep running on the data warehouse. This is a documented Snowflake behavior.
In a model, an input can’t use columns named “MAIN_ID”, “OTHER_ID”, “OTHER_ID_TYPE”, or “VALID_AT” in its ID SQL.
When creating a connection via init command, pressing the Ctrl+C command doesn’t exit the application.
migrate auto jumbles up the order and removes comments.
On Redshift, some experimental commands such as discover do not work.
The command migrate auto migrates siteconfig in your home directory but not any local one.
While working with same type of data in Snowflake and Redshift you might encounter errors where it works on Snowflake but not on Redshift. This is due to the fact that implicit casting of different data types for different function or operator might not be supported on one data warehouse while supported on other.

Version 0.3.1

3 February 2023

This version addresses a crucial defect, so please make sure to update your version. Note that you won’t have to update your schema for this release.

Version 0.3

25 January 2023

We have got a new name! WHT is now called Profile Builder (PB), RS360 is now Profiles. Be sure to check out our newest release that comes with several new features for an enhanced user experience.

What’s New

Migrate - A new command that will enable you to migrate your project to a newer schema. It has two subcommands:
- Manual - You will get to know steps you need to follow to manually migrate the project yourself. It will include both breaking and non-breaking changes.
- Auto - Automatically migrate from one version to another.
We have made a few significant changes to YAML. The changes consist of:
- Bumping schema version from 9 → 18.
- Entityvar (Feature Table) - We have renamed tablevar, tablefeature and feature to entityvar; as they all were adding columns to an entity with nearly identical YAML. A new vars: section of feature table YAML contains list of inputvars and entityvars. Whereas features: field same YAML is a list of entityvar names which should be part of the final output table.
- ID Stitcher is now linked to an entity. As a result, all tables using that entity will use the linked ID Stitcher. Earlier, an ID stitcher was linked to a feature model.
- Some of the terms in yaml spec are changed to make it closer to SQL terminologies. For entityvar and inputvar spec: value → select, filter → where , ref → from. In inputs spec: sql → select.
- Project file has a new key named include_untimed. If set to true, data without timestamps are included when running models. This reduces data errors for timestamp materials. Also, we have deprecated the flag require-time-clean in the run command.
- Id types can now be re-used between entities. In the project file, entities now have a list of id types names, instead of a list of definitions. In the inputs file, a required entity field is added to the ID list that specifies which entity this ID type is being extracted for.
- Now an inputvar can also read from a macro, just like tablevar.
- Global Macros - You can now define macros in a separate YAML file inside your models folder. They can then be used across different models in your project. Thus a macro becomes independent that can be reused across multiple models.
- wht_project.yaml is renamed to pb_project.yaml and ml-features.yaml to profiles.yaml.
Cleanup Materials - A new command that allows you to review all the created materials and then delete them (NOTE: experimental feature).
Discover - A new subcommand discover materials has been added. Using it, you can now discover all the materials associated with a project.
Compile/Run - GIT URL now supports tags. To use, execute the command pb compile -w git@github.com:<orgname>/<repo>/tag/<tag_version>/<folderpath>.

Improvements

Web app - The UI is now more intuitive and user-friendly.
Log tracing is now enabled by default for most commands. Log files are stored in logs/logfile.log of your current working directory. They store upto 10 MB data. Also, the logger file now stores more granular information for easier debugging in case of unexpected errors.
Significant performance improvements in creating ID stitched tables, in case a lot of duplicates are present.
Add extra columns (Hash, SeqNos) to differentiate between entries for commands to discover sources and entities.
When you execute a profile via run command, then the generated SQL gets saved in the output folder.
Added .gitignore file to init project command, to prevent unnecessary files being added to GIT Repo. Such as, .DS_Store, output and logs folders.
Tonnes of changes under the hood.

Bug Fixes

Fixed the bug where window functions were creating multiple rows (duplicates) per main id.
Resolved the bug in inputvars which was doing joins on main_id instead of row id.
Executing the command init profile now inputs values in the same order as on the web app.
Resolved the bug where extra gitcreds[] and warehouse lines were added on overwriting a profile that already existed.
A few redundant parameters were being shown in the validate access command which have been removed.
Removed a couple of redundant subcommands in the init project.

Known Issues:

Warning: While the run command is being executed, canceling it by pressing Ctrl+C doesn’t work as expected. Though it will stop the program’s execution on the CLI, the query will keep running on the data warehouse. This is a documented Snowflake behavior.
In a model, an input can’t use columns named “MAIN_ID”, “OTHER_ID”, “OTHER_ID_TYPE”, or “VALID_AT” in its ID SQL.
When creating a profile via init command, pressing the Ctrl+C command doesn’t exit the application.
Web app doesn’t allow you to select a date older than 30 days.
Migrate auto jumbles up the order and removes comments.

Version 0.2.2

12 November 2022

Our November release is significant as it has several fixes and improvements for an enhanced experience. Check it out and be sure to let us know your feedback.

What’s New

ID Stitcher / Feature Table - You can now define a view as source, in addition to table, in the inputs file. This is particularly of use when you need to support an sql query that’s complex or out of scope for PB. To use it, in your inputs file define the edge_source as view: <view_name> instead of table: <table_name>.
Inputvars - A new identifier which adds temporary helper columns to an input table, for use in calculating a featuretable.
Window Functions - In your model file, you can now add window function support to features, tablevars, tablefeatures and inputvars. Also, you can add filters to features.

Improvements

Schema version 9 makes it more streamlined to define the model. We welcome your feedback for further improvements on this.
Compile command now show errors if the input SQL is buggy.
Discover - subcommands entities and features now show a few more fields.
Discover - Export to CSV works for subcommands and also generates files in the output folder.
Init pb-project - Based on feedback, it now generates a README file and also has simpler YAML files with comments. It should now be easier for our users to create a model and get it running.
Several internal refactorings on how the application works.
Web app - Massive improvements under the hood related to UI elements, preserving state when entering data, showing correct data and validations, and displaying run time in user’s local time zone.

Bug Fixes

Fixed the issue where every time pb run was executed for a feature table, it was adding a new row to the output of pb discover features.
Resolved the bug where error wasn’t shown if an unknown flag was used.
There was an issue generating material tables on a new schema, which has now been resolved.
Bug fix on generating empty SQL files from input models.
Fixed bug where model names with _ in the name would sometimes fail to update the latest view pointer correctly.
Web app - Artifacts list now shows different folders for different runs to isolate them.
Web app - When the PB project is running, the screen now shows correct start timestamp.
Web app - Date filters to find PB runs are now working.
Web app - Scheduling UI is now fully responsive about when the run will take place.
Web app - Resolved the issue where a project would run only once and was then showing error.

Known Issues:

Warning: While the run command is being executed, canceling it by pressing Ctrl+C doesn’t work as expected. Though it will stop the program’s execution on the CLI, the query will keep running on the data warehouse. This is a documented Snowflake behavior.
In a model, an input can’t use columns named “MAIN_ID”, “OTHER_ID”, “OTHER_ID_TYPE”, or “VALID_AT” in its ID SQL.
When creating a profile via init command, pressing the Ctrl+C command doesn’t exit the application.
Logger file generation is disabled at the moment.
Some no-op parameters are shown upon passing the help flag(-h) to validate access command.

Version 0.2

5 October 2022

The September release is our largest update yet. We have added a lot of quality of life improvements and net new features to the PB product line. We plan on releasing even more features in our mid-October release to further improve the usability of the product as well as add additional features that will further help form the core of the product. A substantial amount of the features in this release were based directly off feedback from the first beta testing with external users and internal stakeholders. Please feel free to walk through our newest release. We welcome and encourage all constructive feedback on the product.

What’s New

Feature Table - After encouraging feedback from beta testing of the ID Stitcher, we are feeling more confident about sharing our C360 feature table functionality with beta customers. During testing of this release, we benchmarked ourselves against the feature set that our E-Commerce ML models expect. Many features were implemented successfully. Some needed functionality which could not be pushed through QA gates in this September release. Nevertheless, the feature table YAML is now ready for internal customers to explore.
Web App - We are now ready to share the scheduling functionality within the web app. This will allow the user to schedule, and automatically run PB models from the Rudder backplane. Any artifacts and log files created during the execution of PB projects are also available for the user to explore. This critical functionality will enable users to debug their cloud PB runs.
Validate - A new command, pb validate allows users to run various tests on the project related configurations and validate the privileges associated with the role used for running the project. For example, the subcommand pb validate access does an exhaustive test of required privileges before accessing the warehouse.
Version - This is another new command that provides information on the current version of the app.
Logger - When you execute the compile and run commands, all errors and success messages that were previously only displayed on screen, are now also logged in a file inside the project output folder.
Discover - You can now export the output of the discover command in a CSV file. The ability to discover across all schemas in one’s warehouse is also added.

Improvements

We have made many changes to the way ID Stitcher config is written. We are forming a more complete opinion on the semantic model representation for customer’s data. Entities, IDs, and ID types are now defined in the PB project file. The model file syntax is also more organized and easier to write. To see examples of the new syntax check out the section on Identity stitching or sample files by executing command pb init pb-project. The sample project file also contains include and exclude filters, to illustrate their usage.
In PB command invocation, whenever a file is written, its location is now shown on the console and in log files.
Many enhancements on how errors are handled inside the application.
Massive improvements under the hood.

Bug Fixes

Fixed the issue in ID stitching where it was not picking up singleton components (i.e. the ones with only 1 edge), due to which they were getting skipped in the final output table.
In the init command, not entering any value for target wasn’t setting it to default value as “dev”.
Pressing Ctrl+C wasn’t exiting the application.
The command init profile now appends to an existing profile, instead of overwriting it.
Fixed the issue in discover command where the material table name was being displayed instead of the model name.

Known Issues:

Warning: While the run command is being executed, canceling it by pressing Ctrl+C doesn’t work as expected. Though it will stop the program’s execution on the CLI, the query will keep running on the data warehouse. This is a documented Snowflake behavior.
In a model, an input can’t use columns named “MAIN_ID”, “OTHER_ID”, “OTHER_ID_TYPE”, or “VALID_AT” in its ID SQL.
The web app is not showing a description and last run on the landing page.
In the web app, date filters to find PB runs aren’t working.
In the web app, when the PB project is running, the screen shows an incorrect start timestamp.
Artifacts list changes when a project is running versus when it completes execution. Since all runs on the same Kubernetes pod share the same project folder, we are creating artifacts of different runs under the same parent folder. So, the same folder is currently shown for different runs of the project. In the next release, we will configure different folders for different runs to isolate them.
In case of feature table models, the compile command doesn’t always show error if the input SQL is buggy. Thise error may still be found when the model is run.
When creating a profile via init command, pressing the Ctrl+C command doesn’t exit the application.
Creating a PB Project doesn’t currently include a sample independent ID stitcher. Instead, it is a child model to the generated feature table model.
We are working toward better readability of the logger file. We welcome any feedback here.
The command pb discover features needs to show a few more fields.
Every time pb run is executed for a feature table, it adds a new row to the output of pb discover features. Only one row should appear for each feature.
Export to CSV for the discover command should work for subcommands and also generate files in an output folder.
Some no-op parameters are shown upon passing the help flag(-h) to validate access command.
In some cases, error isn’t shown if an unknown flag is used.
Scheduling UI isn’t sometimes fully responsive about when the run will take place.

The documentation for September release does not completely match with the current release. We are currently working on updating the documentation and will have new versions out soon. Please contact the Data Apps team if you are confused by some deviation.

Version 0.1

18 August 2022

We are now in beta! Please do try out PB and share your feedback with us, so that we can make it better.

What’s New

ID Stitcher - ID Stitching solves the problem of tying different identities together, for the same user across different sessions/devices. With v0.1, we launch PB ID Stitching. It provides an easy to use and powerful interface to specify Id Stitching inputs.
Command Line Interface - Our CLI tool works on Linux, Windows and Mac machines. Using it you can setup a profile having connection to your Database, make a PB project, create SQL from models, run ID stitcher models directly on the Warehouse, and discover all the created models/entities/sources on DW.

Improvements

We have enhanced the speed of Discover and Compile commands, from minutes to a few seconds.
The description of a few commands in Help has been improved.

Bug Fixes

The command for discovering entities wasn’t working, which has now been resolved.
Fixed the bug on init profile command where siteconfig wasn’t getting created on first-time installations.
A few bugs resolved related to output of discover command.

Known Issues:

Warning: While the run command is being executed, please do not cancel it by pressing Ctrl+C. Though it will stop the program’s execution on CLI, the query will keep running on the data warehouse. This is a documented Snowflake behaviour.
Null ID’s in ID stitcher. If first listed Id is null, the entire row may be ignored. That means, results are silently incorrect.
If first listed ID is null, the entire row may be ignored. The first listed ID is assumed to be the key ID. If it is ever null the results may be incorrect.

Was this page helpful?

Glad to hear it! Please tell us how we can improve.

Sorry to hear that. Please tell us how we can improve.

Questions? Contact us by email or on Slack