HFSQL

• find the author, the date and time of an operation performed on a specific row • store the history of a table use (to per- form statistical calculations for example) These operations can be run from a com- mand line, from the WDLog tool or via programming. SQL VIEWS An SQL view is a “virtual data source”, defined via an SQL query. All the SQL views created are kept on the HFSQL server, and they can be reused from the applications accessing the data- base. An application can perform queries on these SQL views. An application can use SQL views in order not to be dependent of the physi- cal organization of data in the database. “Materialized views” are also available. The main difference is that the result of a “materialized view” is physically stored on the server disk. Unlike an “SQL view” which is re-ex- tracted at each call, a “materialized view” stores the data on the drive. BACKUPS Backups are important features of a data- base. You can save the entire server’s content, only the database, or only a selection of tables with or without index. HFSQL supports hot backups, as well as differential backups. A backup is portable, for instance from a Windows server to a Linux server, or from a Client/Server version to a Classic version. Backup configuration A backup can be triggered from the ad- ministration tool, Control Center (instant backup, scheduled backup), or via pro- gramming, directly from the application. The frequency of full backup and differen- tial backup can be specified. For example: 1 full backup every month and a differen- tial backup once a week. The execution of stored procedures before and/or after the backup lets you perform automated processes : send email, copy the backup to a network location, etc... The number of backups to keep can be specified. A backup can be performed “hot”, with- out disconnecting users, transparently, without interfering with the applications. “HOT” ADMINISTRATION A large number of maintenance tasks can be performed live (“hot”), without need- ing to disconnect users, and without inter- fering with their running applications. Applications continue to read and write data during these phases: • Hot “auto modif” DSS • Hot reindexing • Hot automatic optimization of perform- ances • Hot change of password • Backup. FRAME COMPRESSION A frame is a data packet that flows on the network. HFSQL, like all the DBMSs, is using frames to establish communication between the server and client computer. A server’s data transfer speed depends on the travel speed of the frames through the network and on the size of these frames. Frame compression allows you to reduce the size of the packets traveling over the network. In a context of remote communication, frame compression can be very important. The speed for remote connection is im- proved. “BLOB” DATA COMPRESSION “blob” type data (text and binary memos) can be compressed to optimize the space used on the disk. The space used can be significantly re- duced this way. UNALTERABLE TABLE: AN UNMODIFIABLE TABLE An unalterable table is a table in which it is impossible to do anything other than adding lines. It is impossible to modify or delete lines, or to modify the structure of the table. OBSOLETE TABLES & ITEMS: ZOMBIES When a table or an item (column) must no longer be used, but it cannot be deleted from the data description, it is possible to mark them as “zombie” tables or items. It exists, but it must not be used in the new code that is created. GDPR (PERSONAL DATA) HFSQL allows you to abide by the GDPR rules (see page 16). LINK WITH OTHER DATABASES HFSQL can be used simultaneously with other databases. Most IT departments use several heterogeneous databases. HFSQL also lets you exchange data with other databases. HFSQL • www.windev.com 9