Release Notes

Bug fixes

With Celonis Process Mining 4.7.3, we are deprecating parts of the “optimized load” functionality (Data Model → Loading → Load column).

Remaining functionality:

The user can manually select which columns of a certain table are included/excluded from a data model load.

Deprecated functionality:

• The application no longer tracks the access count and last access dates for columns
• The “Load only used columns” functionality is no longer available. This functionality was used to automatically include columns that were not available but accessed by a user into the next load

2. MS SQL Configuration Store - JDBC Driver

In case the application is configured to use the Celonis Configuration Store based on MS SQL in config-custom.properties, a manual  update has to be applied:

Validating the usage of MS SQL

• Navigate to <installDir>/config-custom.properties
• Check if this exact property exists within your config-custom.properties file. In case the property is commented (#) the manual update does not have to be applied:

database.dialect=org.hibernate.dialect.SQLServer2008Dialect

Please refer to the Celonis Update Guide 1.15 for more information. Not applying this configuration can lead to MS SQL connections failing.

How to update the property

1. Navigate to <installDir>/config-custom.properties
2. Replace the old value org.hibernate.dialect.SQLServer2008Dialect of the property database.dialect with the following value:

de.celonis.pm.utils.engine.CPMSQLServer2012Dialect

3. Microsoft JDBC Driver for SQL Server Authentication Library for Windows

We also removed the mssql-jdbc_auth-<version>.x<64/86>.dll (Microsoft JDBC Driver for SQL Server Authentication Library for Windows) from the embedded dependencies. In case the application is running on Windows and the SQL Server Authentication (indicated by integratedSecurity=true in the JDBC string) is used to connect to the Celonis Configuration Store or the Analytics Database, the library has to be deployed manually.

Please refer to the Celonis Update Guide 1.15 for more information. Not applying this configuration can lead to MS SQL connections failing.

How to deploy the driver

1. Stop the application
2. Download the respective driver version based on your SQL Server version and compatibility from here. Please follow the Microsoft recommendations in case of any questions: here.
3. Place the library in <installDir>\jre\bin
4. Start the application

4. Updated log file structure & log rotation

With version 4.7.3, we revamped the full log file structure on the application server. In addition to that, we enabled log rotation by default for the main Celonis Frontend and Compute Compute service logs.

Please refer to the Celonis Update Guide 1.15 for more information. 

5. Custom logging configuration

The custom logging configuration was updated with version 4.7.3. It was previously configured in <installDir>config-custom.properties via the property logging.config=. This property is no longer supported. 

• If custom logging was configured previously, please migrate the configuration to the approach (see the next slide).
• If scheduled OS logrotate was configured, please deactivate it and use the default CPM4 log rotation. Alternatively, you can apply a custom logging configuration as highlighted on the next slide.

Please refer to the Celonis Update Guide 1.14 for more information. Not applying this update will lead to the application not picking up the previous custom logging configuration.

How to apply a custom logging configuration

• Refer a custom logging config file using the JVM property -Dlogging.config in <installDir>/cpm.user.vmoptions
• Refer a custom logging configuration in the file application.properties#logging.config residing in the same directory as the application jar.

1. Custom dimension

The Conformance Checker component can now be configured to use a custom dimension in two places (requires being in edit mode):

• During the setup when selecting “mine process model”
• In the component settings

2. Improved violation list

The violation list is now paginated for better usability and performance.

Additionally, violations can now be added to the allowlist directly, without opening the detail screen.

1

2.

Previously, the color mapping of analysis components was limited to numeric values. Therefore, we added the option to map null values to a specific background or font color.

We added more information to the “Scheduled loads” overview (System Settings → Data loads → Scheduled loads):

• Data model ID
• Load ID
• Load schedule/partial load schedule
• Schedule configuration

1. Storage metrics

The application now logs the available and used storage of config.properties#filesystem.writableroot= (once per hour).

2. Data Model Load Pool metrics:

The application now logs information about the Data Model Load Pool configuration (config-custom.properties#model.max.parallel) and workload (once per hour).

Examples:

<yyyy-MM-dd'T'HH:mm:ss.SSSXXX> INFO ResourceMonitor - [STORAGE (writable root)] available 64 GB, used 32 GB (50.00%)

<yyyy-MM-dd'T'HH:mm:ss.SSSXXX> INFO ResourceMonitor - [Data Model Load Pool] current pool size = 3, maximum pool size = 6, active tasks = 0, completed tasks = 3, queued tasks =