IBM Support

cannot "memget" the "memputted" member with memidnum ÀBÇD1234 as is



When storing records from different source systems in Virtual MDM, those records contain a primary key (memrecno - a long) defined by MDM and also a foreign key from the source system (memidnum - a string). Both of these reside in the header for each record (MemHead). It is possible to query records by either primary or foreign key. If the forein key contains non-latin characters (eg ÀBÇD123), attempts to query the record by this field fail and return no data. The record can still be retrieved by query with the primary key. Typically, key fields don’t contain text as such and are not likely to hold such characters – they are more likely to contain numeric or simple latin alphabet characters. Whilst this is relatively unlikely to impact users, we should still investigate the query operation to work out what the problem is and consider for an early fixpack post E release. Note also that there was discussion that this may actually be caused by problems with the test framework.

Resolving The Problem

If customers do have source systems that have primary keys that contain these characters, then converting them to Unicode format will allow them to be stored and retrieved as foreign keys in MDM. For the example above the key could be stored as \u00c0\u0042\u00c7\u0044\u0031\u0032\u0033.

[{"Product":{"code":"SSWSR9","label":"IBM InfoSphere Master Data Management"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Component":"--","Platform":[{"code":"PF002","label":"AIX"},{"code":"PF016","label":"Linux"},{"code":"PF033","label":"Windows"}],"Version":"","Edition":"","Line of Business":{"code":"LOB10","label":"Data and AI"}}]

Document Information

Modified date:
27 April 2022