There are many new enhancements in DB2 11 for z/OS, and some of the most important ones are the ones that help data management handle the schema chaos within application development. Since the colleges are not truly teaching data management, modeling, or standard design normalization practices anymore, development constantly changes the data elements within the schema.
DB2 11 for z/OS helps address this schema chaos through three new enhancements that help all of us keep up with the database changes, minimize the related downtime, and handle new data types.
First, within DB2 11 for z/OS it is now possible to DROP a column from a DB2 table. While there are many restrictions around DROPing a column from a table the ALTER table-name DROP COLUMN statement is now available in DB2 11. This DROP column capability provides the ability to remove obsolete data elements from tables and reduce the row size of your DB2 tables.
Schema chaos changes can also be addressed through all the other DB2 10 for z/OS ALTER enhancements that can convert columns to different data type definitions provide further flexibility to ALTER a table and its column definitions to anything that the application needs. The new DB2 11 for z/OS simple ALTER DROP COLUMN statement with subsequent database table reorganization easily incorporates the new table definition. If the database table changes are done online for the development environment, the developers can even continue working
The next way DB2 11 for z/OS helps us all address the application development schema chaos is through ALTER Partition Limit Keys functionality. This helps all of us expand or contract the partitioning key limits defined within our large table definitions. Again, of course, there are many table space and table definition restrictions that should be researched before table creation to make sure your partitioning scheme can be ALTERed later, but the ability to change the limit keys can help your design maximize the number of rows within partitions.
For example if an increasing customer number, order number or other key or keys combination is used within the limit keys definition, the number of rows could become too big or too small as the business activity grows or shrinks. The ability to ALTER the Partition Limit Keys functionality provides the ability to easily expand or contract a partition to fit the data. This can make a big difference if your design needs to consolidate the underlying partition data sets. In this era of Big Data and longer data retention time requirements, it is a great option to consolidate older data partitions together or improve your partitioning to handle more data overall.
These older consolidated partition data sets could be pushed into an IBM DB2 Analytics Accelerator (IDAA) solution using the high performance super saver (HPSS) option that takes the data out of the DB2 table partition and pushes it into the IDAA box. So using the ALTER of the partitioning limit keys opens up a wide variety of options for optimizing data storage and performance of queries in your database, especially your Big Data tables.
Another way to address the schema chaos is to embrace it through the new DB2 11 for z/OS JSON data store. JSON, JavaScript Object Notation, is the latest Java interface for data at rest. JSON is a simple self-describing text based open standard that is optimized for high speed ingest activities. Since JSON is very easy to understand and self-describing as a tag and value pair, it is quickly changed for rapid application development. Putting JSON inside of DB2 11 provides JSON with the ACID (Atomicity, Consistency, Isolation, Durability) database concepts of “commits,” logging, indexing, and SQL access.
JSON is replacing XML as developers choose the new easy JSON interface for all types of applications. Web applications, mobile phone applications as well as C, Java and other new developer application languages are embracing JSON as their API to the retrieve data. HTML5, AJAX, and JSON are the new combination for all types of applications. DB2 11 for z/OS with its new JSON API and data store handles the schema chaos easily.
Leave a Reply