180 likes | 245 Views
Design for the Future. Equal Format Data Systems. The Future of Data. For the DBA. Relational Design Methods: Efficiency through Normalization. Reducing empty space and increasing flexibility by clustering around entity types and attribute groups.
E N D
Design for the Future Equal Format Data Systems The Future of Data For the DBA
Relational Design Methods: Efficiency through Normalization • Reducing empty space and increasing flexibility by clustering around entity types and attribute groups
Problems of Relational Table Design: Entities and Relationships Simple General-Purpose Relationship Table • Real world entities have many possible simple relationships.
Problems of Relational Table Design: Entities and Relationships • Multiple entity tables. • N2 possible relationships.
Problems of Relational Table Design: Entities and Attributes Old Entity Table • Putting everything with options into the same entity table leads to lots of unused space.
Step 1 Equal Format Table Design: Convert Attributes to Relations • Attributes are equivalent to relations. • Relationship Types are also entities.
Step 2 Equal Format Table Design: Convert Model from E-R to Semantic Triplet Sentence Table Phrase Table • Change the Entities to Phrases. • Change the Relations to Semantic Triplet Sentences.
Step 3 Equal Format Table Design: Different Types of Phrases Phrase Table • Allows any data type in any “cell”. • Common value types are pooled and indexed. ValueFloatTable
Step 4 Equal Format Table Design: Sentences about Sentences Phrase Table • Used for Project management, Security filtering • Used for metadata in the style of “(S1) was captured by (Device19)”.
Are we done? Not yet. Linguistically speaking: Triplets are baby talk Complex Relationship Table Example • This table requires either more phrases in a sentence or more sentences.
Step 5 Equal Format Table Design: Normalize the Sentence Table • Allows more phrases per sentence, more phrases per syntax position, and more syntax positions (metadata). • Now you can select “tables” and “columns” the same way as rows.
Final Step 6 Equal Format Table Design: Full Sentence Diagrams Value Sentence Table • The sentence table can now hold any complex relationship.
Equal Format Databases: A Different Type of Container This Old Table • The old design pattern - meaning is dependent on the row and column positions in each table. • The newdesign pattern - pieces of Equal Format Data. Tables of Equal Format Data
Equal Format Databases: Scalable System Designs • Common container types reduce N2 complexity in coding, point-to-point conversions and messaging. • Development is de-linked from structural dependency. • Atomic data can be modified non-destructively more efficiently with less locking. • Nondestructive storage can maintain a running snapshot • Reusable structures encourage reusable code and common dictionaries. • Atomic data enables multiple inheritance
Equal Format Databases: Tradeoffs and Design Guidelines • There are many ways to say it – keep it simple. • Try to be less ambiguous and use stricter and simpler syntax. • It’s easier to change the vocabulary than change the structure. • More wordiness uses space but adds meaning. • Compared to traditional design, it uses more space per cell when nonempty, but no space per cell when empty. • Repeat the container as multiple “subsystems” to tune the structure
Equal Format Data: Additional Applications and Uses • Master Data Storage • Messaging Systems • Data Warehousing, Data Marts, and Data Mining • Security, Auditing, Quality Control • Merging Legacy Information • Facility Data and PLM Data • Engineering Project Management • Product Catalog Data • Development Infrastructure
Datura, LLCthe Leaders in Equal Format Databases www.datura-llc.com Datura, LLC PO Box 498 Howell, MI 48844-0498 USA sales@datura-llc.com +1 (734) 878-0109 voice +1 (734) 878-0684 fax