What can be migrated to IBM Cloud Pak for Data (IBM Knowledge Catalog)

Check which data you can migrate from IBM InfoSphere Information Server to Cloud Pak for Data 5.4.

For information about the replacement features and where to find your data after the migration, see Where you can find the migrated data.

Depending on your operating system, you can migrate data that is stored in an InfoSphere Information Server metadata repository on one of these database types:
  • IBM Db2
  • Microsoft SQL Server
  • Oracle
For more information, see Product versions supported for migration.

Users, user groups, and permissions

Migrating users, user groups, and permissions is a separate migration step and can be done only if the InfoSphere Information Server system is configured to use an LDAP user registry and both systems are connected to the same LDAP server.

For more information, see these topics:

Data connections created for automated discovery or use in data quality projects

Data that can be migrated Data that won't be migrated
Connections for the following connectors:
  • Amazon S3
  • Apache HBase Connector (to connections of the type Generic JDBC)
  • Azure Datalake Storage Connector
  • BigQuery Connector
  • File Connector - Engine tier (to connections of the type Generic JDBC)
  • File Connector - HDFS
  • Cloud Object Storage Connector
  • Google Cloud Storage Connector
  • Greenplum Connector
  • Hive Connector
  • IBM Cognos TM1 Connector (to connections of the type IBM Planning Analytics)

    You must manually update the SSL certificate for the migrated connection.

  • IBM Db2 for z/OS

    You must complete some post-migration steps for connections of this type. See Configuring IBM Db2 for z/OS connections.

  • IBM InfoSphere DB2 Connector
  • IBM Netezza Connector
  • JDBC Connector
  • ODBC Connector (to connections of the type Generic JDBC for unsupported data sources)
  • Oracle Connector
  • Redshift Connector
  • Snowflake Connector
  • SQL Server
  • Teradata Connector

Kerberos connections and any data added by using such connection are migrated but the migrated connection cannot be used.

Also, the following information is not checked during the export:
  • Whether the connection password is active or expired
  • Whether a platform connection was deleted after it was synced to the legacy metadata import component
 None.

Import areas

Data that can be migrated Data that won't be migrated
You can migrate import areas with connections for the following connectors:
  • Amazon S3
  • Apache HBase Connector (to connections of the type Generic JDBC)
  • Azure Datalake Storage Connector
  • BigQuery Connector
  • Cloud Object Storage Connector
  • File Connector - Engine tier (to connections of the type Generic JDBC)
  • File Connector - HDFS
  • Google Cloud Storage Connector
  • Greenplum Connector
  • Hive Connector
  • IBM Cognos TM1 Connector (to connections of the type IBM Planning Analytics)

    You must manually update the SSL certificate for the migrated connection.

  • IBM Db2 for z/OS

    You must complete some post-migration steps for connections of this type. See Configuring IBM Db2 for z/OS connections.

  • IBM InfoSphere DB2 Connector
  • IBM Netezza Connector
  • JDBC Connector
  • ODBC Connector (to connections of the type Generic JDBC for supported data sources)
  • Oracle Connector
  • Redshift Connector
  • Snowflake Connector
  • SQL Server
  • Teradata Connector
You can also migrate import areas with connections for the following MITI bridges (only connections migrate successfully):
  • IBM Cognos
  • Tableau Server

Use of migrated content is limited to the connections.

Import areas include import area details, connection details and import parameter details. Imported metadata is not migrated and is not supported for migration. After the import area is migrated, you can perform a re-import to get the source metadata.

 You cannot migrate import areas with connections for any MITI bridges other than IBM Cognos and Tableau Server.

You cannot migrate assets from a staging area.

Information assets

Data that can be migrated Data that won't be migrated
You can migrate the following types of assets:
  • Database tables and data files are directly migrated as data assets, where data files cannot have more than one data file record.
  • Database schemas, data file folders, and data file fields are migrated as properties of data assets.
  • Host and database are migrated as properties of a data connection asset.
  • Aliases of database tables are migrated as data assets with the table type ALIAS.
The following properties can be migrated:
  • Custom properties and custom relationships
  • Implements rules
  • Governed by rules
  • Notes
  • Labels
  • Assigned to terms
  • Candidate and foreign keys
  • Stewards
  • Column and data quality analysis results
You cannot migrate these items:
  • Indexes
  • Data files that have more than one data file record
  • For data files, the IGC Alias (Business Name) property

Logical and physical data models

Data that can be migrated Data that won't be migrated
You can migrate the following assets and properties:
Logical data models
  • Logical data model
  • Logical entity
  • Logical relationship
  • Entity attribute
  • Attribute implemented by data field
  • Custom properties
  • Custom relationships of type Steward
  • Notes
  • Terms
Physical data models
  • Physical data model
  • Design table
  • Design view
  • Design column
  • Default properties
  • Physical constraints
  • Custom properties
  • Custom relationships of type Steward
  • Notes
  • Terms
 The following data cannot be migrated and is no longer available in Cloud Pak for Data 5.4.
Logical data models
  • Relationships for logical data models implemented by logical data models, a data schema, or a model package
  • Relationships for a logical entity implemented by data collections
  • The IGC Alias (Business Name) property
  • Governance rules bound to logical data model assets
Physical data models
  • Relationships for physical data models implemented by a data schema, data files, or a model package
  • Primary keys
  • The IGC Alias (Business Name) property
  • Stored procedures
  • Governance rules bound to physical data model assets

Business intelligence assets and asset types

Data that can be migrated Data that won't be migrated
You can migrate these types of business intelligence assets:
  • BI Report
  • BI Report Query
  • BI Report Query Item
You can migrate the following information for business intelligence assets:
  • Asset name
  • Short description
  • Long description
  • Context relationships
  • Labels
  • Assigned to terms
  • Governed by rules
  • Implements rules
  • Alias
  • In collections (migrated as tags)
  • Notes
  • Custom properties
  • Stewards
  • Images
You can migrate the following information for business intelligence assets, but it will be available only in the source_system_history and will not appear in the UI:
  • Created by
  • Created on
  • Last modified on
 None.

Extended data sources

Data that can be migrated Data that won't be migrated
You can migrate the following information for extended data source assets:
  • Asset name
  • Short description
  • Long description
  • Context relationships
  • Labels
  • Assigned to terms
  • Governed by rules
  • Implements rules
  • Alias
  • In collections (migrated as tags)
  • Notes
  • Custom properties
  • Stewards
  • Images
You can migrate the following information for extended data source assets, but it will be available only in the source_system_history and will not appear in the UI:
  • Created by
  • Created on
  • Last modified on
 None.

Extension mapping documents

Data that can be migrated Data that won't be migrated
You can migrate the following information for extension mapping documents:
  • Asset name
  • Short description
  • Long description
  • Context relationships
  • Labels
  • Assigned to terms
  • Governed by rules
  • Implements rules
  • Alias
  • In collections (migrated as tags)
  • Notes
  • Custom properties
  • Stewards
  • Images
You can migrate the following information for extension mapping assets, but it will be available only in the source_system_history and will not appear in the UI:
  • Created by
  • Created on
  • Last modified on
 None.

OpenIGC assets and asset types

Data that can be migrated Data that won't be migrated
You can migrate the following information for OpenIGC asset types:
  • Asset type name
  • Bundle property definitions
  • Bundle properties organization (sections)
  • Properties and sections inheritance behavior
  • Bundle types containment definitions
  • Translations
You can migrate the following information for OpenIGC assets:
  • Asset name
  • Short description
  • Long description
  • Context relationships
  • All bundle property values (single value, multi-value, (not)editable, String, LongText, Date, Integer, Double, Boolean)
  • Labels
  • Assigned to terms
  • Governed by rules
  • Implements rules
  • Alias
  • In collections (migrated as tags)
  • Notes
  • Custom properties
  • Stewards
You can migrate the following information for OpenIGC assets, but it will be available only in the source_system_history and will not appear in the UI:
  • Created by
  • Created on
  • Last modified on
Limitations
The following limitations apply to OpenIGC assets:
  • Changes to custom types that were representing supertypes in the source component are not inherited.
  • OpenIGC asset types that are imported as a part of the same bundle are not organized with the bundleId identifier in the target catalog.
  • The concepts of bundle family and bundle hierarchy are not available in the target catalog.
  • Source and target components support different language sets. Not all translations will be available in the target catalog.
  • The plural form of type names is not supported in the target catalog.
  • The property expandableInLineage is not supported in the target catalog.
 You cannot migrate OpenIGC type icons.

Lineage information

For importing lineage information to IBM Cloud Pak for Data, MANTA Automated Data Lineage for IBM Cloud Pak for Data must be installed and an active license must be available.

Data that can be migrated Data that won't be migrated
Lineage for these types of assets:
  • BI Reports
  • Extension Mapping Documents
  • OpenIGC assets with lineage flows
 You cannot migrate these items:
  • Lineage configuration, templates, and filters
  • Lineage for BI Models
  • Lineage for Data Connection Mappings
  • Lineage for Database Schema Identity Mappings
  • Lineage for Extension Data Sources
  • Lineage for InfoSphere DataStage jobs
    Tip: You can manually migrate such jobs as described in Migrating DataStage jobs and re-create the lineage flows by using MANTA Automated Data Lineage.
  • Lineage for Mapping Specifications
  • Lineage for MDM Models
  • Operational lineage

Glossary asset types

Only published versions of glossary assets are migrated (terms, categories, information governance rules, and information governance policies). In the target catalog, they are also imported as the published artifacts. .

Data that can be migrated Data that won't be migrated
Categories
The following properties are migrated:
  • Name
  • Parent Category
  • Short Description
  • Long Description
  • Labels (migrated as tags)
  • Subcategories
  • Contains Business Terms (migrated as Parent Category relationship)
  • Collections (migrated as tags)
  • Custom attribute values
  • Notes (migrated as a comments in the activity log)
The following properties are not migrated:
  • Stewards
  • Assigned to Terms
  • References Business Terms
Data classes
The following properties are migrated:
  • Name
  • Short Description
  • Long Description
  • Example (migrated as part of the description)
  • Labels
  • Stewards
  • Enabled
  • Type
  • Minimum Data Length
  • Maximum Data Length
  • Provider
  • Priority
  • Scope
  • Threshold
  • Assigned to Terms
  • Implements Rules
  • Collections (migrated as tags)
  • Custom attribute values
  • Notes (migrated as a comments in the activity log)
Special considerations:
  • For data classes of type Regex:

    Cloud Pak for Data supports only one regular expression for a data class. If you have data classes with an additional regular expression defined, you must split these data classes in your source system so that each has only one regular expression. If you don’t split such data classes, only the main regular expression is migrated.

  • For data classes of type Java:

    Data classes that use a custom Java class are not migrated.

 None.
Information governance rules
The following properties are migrated:
  • Name
  • Short Description
  • Long Description
  • Labels
  • Stewards
  • Related Rules
  • Governs Assets
  • Referencing Policies (migrated as parent policies)
  • Collections (migrated as tags)
  • Custom attribute values
  • Notes (migrated as a comment in the activity log)
The following property is not migrated:
  • Implemented By Assets
Information governance policies
The following properties are migrated:
  • Name
  • Short Description
  • Long Description
  • Labels
  • Stewards
  • Subpolicies
  • Information Governance Rules
  • Collections (migrated as tags)
  • Custom attribute values
  • Notes (migrated as a comment in the activity log)
None.
Labels None.
Terms
The following properties are migrated:
  • Name
  • Parent Category
  • Short Description
  • Long Description
  • Referencing Categories
  • Labels
  • Stewards
  • Governed by Rules
  • Abbreviation
  • Additional Abbreviation
  • Example (migrated as part of the description)
  • Usage (migrated as part of the description)
  • Is a Type Of
  • Has Types
  • Is Of
  • Has A
  • Synonyms
  • Preferred Synonym
  • Related Terms
  • Assigned Terms
  • Assigned to Terms
  • Assigned Assets
  • Collections (migrated as tags)
  • Custom attribute values
  • Notes (migrated as a comment in the activity log)

In addition, the term history is migrated.
The following properties are not migrated:
  • Status (Candidate, Accepted, Standard, Deprecated)
  • Is Modifier
  • Type
  • Replaces
  • Replaced By
The following information is not migrated:
  • Term history for the relationships Replaces and Replaced By
  • The development log

Automated discovery jobs and results

Data that can be migrated Data that won't be migrated
All information for automated discovery jobs and results is migrated including these properties:
  • Confidence scores of term assignments
  • Custom properties
  • Stewards
  • Suggested terms
 None.

Data quality projects

Data that can be migrated Data that won't be migrated
In addition to automated discovery jobs and results and the individual assets in a data quality project, the following items are migrated:
  • Column and data quality analysis results
  • Data rule definitions, data rules, and rule sets
  • Global logical variables bound to constants
  • Notes that were added to data assets, columns, or rule-related assets
  • Quality rules
  • Relationship analysis results: primary key analysis, foreign key analysis, overlap analysis.
  • Relationships between assets in data quality projects and the following governance artifacts:
    • Business terms
    • Governance rules
    • Stewards
  • Rule history and results
  • Rule schedules
  • SQL virtual tables

    By default, SQL virtual tables are migrated as SQL-based data assets. You can set an export option to have them migrated as SQL-based data quality rules. See Optional export parameters.

  • Terms, notes, labels, stewards, custom attributes, and Implements rules and Governed by rules relationships that are associated with rule definitions, rules, rule sets
After migration, the same connectors as before are supported for data quality rules. 		The following data in data quality projects cannot be migrated and is no longer available in Cloud Pak for Data 5.4:
  • Automation rules.
  • Data quality rule benchmarks.
  • By default, virtual tables that are not built based on SQL statements and virtual columns in data quality projects. Any assets that use such virtual tables or virtual columns, such as data rules, will not be migrated.

    You can force migration of such data rules by setting an export option. See Optional export parameters. Before you can work with the migrated rules, you must reconfigure the bindings by using the IBM Knowledge Catalog API: Update data quality rule.

Customizations

Data that can be migrated Data that won't be migrated
Custom property and relationship definitions, and any values for those custom properties or relationships. Lineage filters and templates cannot be migrated and are no longer available in Cloud Pak for Data 5.4.

Other data that you can migrate

You can migrate this content:
  • Collections and labels are migrated as tags in the target catalog. Tags are plain text values. Therefore, attributes of collections and labels, such as descriptions and notes, are not retained.

Data that you cannot migrate

In Cloud Pak for Data 5.4, you cannot migrate the following data. In some cases, you can re-create the data manually.
  • Analysis results: natural key analysis.
  • Unstructured data sources (IBM StoredIQ assets).
  • Data Science assets (IBM Data Science Experience Local assets).