SR-A4310 · Issue 212995
Fixed locking on node level declare pages
Resolved in Pega Version 7.2
In a scenario where data is being loaded from external system using an activity, the data page should be loaded only once for the first requester and all following requestors should refer to the loaded data page until it is expired. However, when multiple users accessed the data page simultaneously, the data page was being loaded for each one. This was traced to a timing issue with locking, and ReadOnlyDataPageHandler->loadDataPage() has been updated to insure node level data pages do the 'put' operation before releasing a lock.
SR-A10662 · Issue 220678
NVARCHAR2 support added for Oracle
Resolved in Pega Version 7.2
NVARCHAR2 support has been added for Oracle to handle Japanese and other special characters in Report Definition columns.
SR-A392 · Issue 204598
Updated SQLgenerator to avoid creating unneeded implicit DBs
Resolved in Pega Version 7.2
There is a known bug with Oracle JDBC drivers with handling synonyms when using getMetaData to get the database table details. For 7.1 installations/upgrades to split schema, if the customer has synonyms on the database for totally unrelated PRPC schemas it can cause issues with the upgrade. There were various ways to work around the issues (creating private synonyms for all of the tables in the new schema, upgrading to a single schema and then splitting, etc.) but to ease this process the system's SQLGenerator now has a check for zOS specific logic in order to ensure the generic table creation ddl will not be generated when it's not needed, as this was causing unnecessary implicit databases to be created.
SR-A9070 · Issue 221177
Updated caches for Data-Admin-DB class groups
Resolved in Pega Version 7.2
Data-Admin-DB-Table, Data-Admin-DB-Name, and Data-Admin-DB-Class Group were not receiving updated cache entries for newly exposed columns. This was an error in the updated database triggers engine code, and has been corrected.
SR-A11245 · Issue 222247
Designer Studio query/filter matching improved
Resolved in Pega Version 7.2
The activity to fetch instance lists in Designer Studio uses a query which performs an UPPER() operation against the pzInsKey column of the target rule table, but the Ignorecase option was being used for the pzInskey filter. The filter has been modified to use toUppercase edit input for filter value.
SR-A11421 · Issue 223667
nvarchar support corrected for MS/SQL
Resolved in Pega Version 7.2
The datatype nvarchar (used for Unicode encoded strings) in MS/SQL was incorrectly reported as unsupported. The data type mapping has been reviewed and updated to correct this.
SR-A2733 · Issue 221219
Migration process handles underlying varchar CHAR
Resolved in Pega Version 7.2
Handling for CHAR length semantics has been added for imports of rule schema to support Oracle varchar2 columns, and the import/export will correctly deal with the circumstanced columns of CASETYPEVERSION and PXREFOBJECTCLASS.
SR-A11286 · Issue 224545
Drop Index SQL updated to remove duplicates
Resolved in Pega Version 7.2
If a query was used in SchemaManager to get all of the covered indexes for the table, firing the same query in the DB to generate XML resulted in duplicate indexes and column names. This was traced to the Drop Index SQL being generated twice, and the query to get the covered index columns from the DB has been changed to remove the duplicates.
SR-A11286 · Issue 223221
Drop Index SQL updated to remove duplicates
Resolved in Pega Version 7.2
If a query was used in SchemaManager to get all of the covered indexes for the table, firing the same query in the DB to generate XML resulted in duplicate indexes and column names. This was traced to the Drop Index SQL being generated twice, and the query to get the covered index columns from the DB has been changed to remove the duplicates.
SR-A11637 · Issue 228995
Newlines encoded for CSV import
Resolved in Pega Version 7.2
A new line character within the csv fields was treated incorrectly and parsed as an end of a row in csv during import. To avoid this, the parsing logic has been changed to handle new line characters within the fields.