首页 » Big Data and AI » Data Analytics » 正文

Experts Predict Trends for Data Architecture in 2019

From AI and machine learning to data discovery and real-time analytics, creating a strong data architecture strategy is critical to supporting an organization’s data-driven goals.

Greater speed, flexibility, and scalability are common wish-list items, alongside smarter data governance and security capabilities.

Many new technologies and approaches have come to the forefront of data architecture discussions, including data lakes, in-memory databases and engines like Spark and cloud services of all shapes and sizes.

DBTA recently held a roundtable webinar with Michael Coburn, product manager, PMM, Percona; Tyler Mitchell, senior product marketing manager, Couchbase; and Max Neunhöffer, senior developer and architect, ArangoDB, who discussed the top trends in modern data architecture for 2019.

Cloud Migration is on the rise, with 34% of DBTA subscribers anticipate moving a database to the cloud this year, Coburn said. Other trends for 2019 year include:

  • Data Governance is a huge challenge for big data projects – indicated by 65% of DBTA subscribers
  • Data Integration is also a huge challenge for big data projects – indicated by 59% of DBTA subscribers
  • Database Performance is a growing issue – 64% of DBTA subscribers are spending more time on it
  • Data Lakes are moving to the mainstream – 37% of DBTA subscribers use and 20% are considering it
  • NoSQL adoption continues to climb – 27% of DBTA subscribers anticipate adopting a NoSQL database this year
  • Machine Learning is catching on – 20% of DBTA subscribers anticipate adopting it this year
  • Spark is making inroads – 15% of DBTA subscribers anticipate adopting it this year

The traditional approach to data architecture is evolving as monolithic, relational database management systems become increasingly expensive and cumbersome, he explained. Many of these structures lack the ability to respond to change. Sophisticated users expect sophisticated solutions and have no tolerance for slow-loading systems or data interruption.

Performance monitoring can help, Coburn said. Putting data in one place is not sufficient to achieve the vision of a data-driven organization. To benefit from a shared data asset, the proper interfaces need to be made easily available. This could take the form of an OLAP interface for business intelligence, a SQL interface for data analysts, a real-time API for targeting systems or the R language for data scientists.

Mitchell foresees several trends for 2019 including more integration, more scalability, and more hybrid cloud manageability options.

While Neunhöffer predicts several other changes for 2019 such as a focus on multi-model databases, an interest in distributed transactions, and better cloud orchestration.

“Deploying a distributed application must be as easy as installing an app on your phone,” Neunhöffer said.

 

获取更多资讯,欢迎关注微信公众号: atSting