Traditional SQL databases and data warehouses have been around for decades and have been doing really well. However, Big Data NoSQL changed the game with big returns for big investments. Concepts like Edge Analytics have been game changers “uberizing” foundational changes in business models.
I recently attended a conference hosted by Snowflake, a data warehouse vendor revolutionizing the cloud data warehouse industry. Interestingly, data warehousing is changing from traditional star and snowflake models (Kimball and Inmon) to Big Data warehousing, and now to comprehensive cloud data warehousing. While quite a few of my clients say, “cloud is cloudy”, companies like Salesforce, Microsoft, and Amazon need little justification for moving to the cloud with their market share.
Cloud data warehouses are revolutionizing the industry with the concept of NewSQL, which uses traditional SQL with elasticity like NoSQL. After gaining experience with these options, here are my views on the databases:
SQL
Pros:
- Proven database technologies with standard ANSI SQL support
- Ad-hoc querying (business and IT)
Cons:
- Significant issues with scalability
- Complex tuning techniques
NoSQL
Pros:
- 99% availability and significant scalability
- Optimized to support structured, semi-structured, and unstructured data
Cons:
- Evolving programming languages with significant custom coding
- Lacks ACID (Atomicity, Consistency, Isolation, Durability) database transactions
NewSQL
Pros:
- Easy to use scalability and availability eliminating DBA needs
- In memory analytics utilizing Big Data infrastructure
- Standard ANSI SQL features providing easy to use interactions for business and IT users
Cons:
- A combination of traditional SQL and Big Data capabilities offers significant challenges on data governance
- Giving business capabilities to process unlimited data without “teaching them to fish” offers challenges in operations
- Concerns on security at an enterprise level
