In 2020, database management will be a mosaic of old and new technologies. Many people have already implemented relational databases or data warehouses — 86.55 percent — according to the recent DATAVERSITY® Trends in Data Management Report. It will not be hard to see why data warehouses will continue heavy usage in 2020. They integrate transactional and operational data and provide valuable business intelligence (BI) through reporting and analytics.
However, data warehouses will need to be upgraded and complemented with newer technologies for better and quicker database management. Newer options, in 2020, include cloud computing, graph databases, machine learning, and augmented Data Management. Business will need to balance and integrate these old and new database management tools to stay in business and remain competitive during a dynamic year.
Businesses will face conflicting database management priorities in 2020. First, economic and political uncertainties will flatten IT spending, pressuring companies to minimize expenses and take fewer risks with newer technologies.
At the same time, database systems will have to handle huge volumes taking up more computing power. The Internet of Things (IoT) alone “will create about 847 Zettabytes per year by 2021, up from 218 ZB per year in 2016.” Plus, database systems need to work faster, thanks to emerging innovations like quantum computing and machine learning.
Add to this a major drive to improve Data Quality, as only 31 percent of organizations trust their capability to meet digital challenges such as machine learning. Security breaches will quickly increase, costing $5 trillion.
Finally, California’s Consumer Privacy Act (CCPA) takes effect at the start of 2020, adding to the need for database compliance. The combination of all these factors means that companies will need adaptable, secure database systems and simple database management tools and processes.
To meet the goal of simple and adaptable database management, companies will:
- Optimize data warehouses
- Leverage cloud computing
- Include graphs databases
- Consider multi-database systems (MDDBMS)
- Provide machine learning infrastructure
- Augment Data Management
Optimize Data Warehouses
Data warehouses or relational databases will remain a main database management staple over the next year. Relational Database Management Systems (RDBMS) will continue to make up 80 percent of the total operational database marketplace, according to IDC. New applications and projects — 70 percent — will also use relational databases.
RDBMS will need to integrate and work with other older and newer database systems to solve complex business problems and leverage digital tools like machine learning. This will be a database management challenge as traditional relational databases, developed for a static and centralized context, need to function in tandem with decentralized and more complex database systems, like NoSQL databases.
To adapt and simplify their RDBMS, database managers and administrators will migrate these systems to the cloud, if they have not done so already. DATAVERSITY found about 46 percent of participants in a study plan on using cloud-based relational databases in the next year or two.
Cloud computing lowers costs and improves scalability, making it an attractive and popular option. However, other technologies available in 2020 will gain consideration too. Examples include: distributed systems, which manage data over a variety of nodes, spreading out the database workload; graphic processing units (GPUs), chips that power and speed relational databases; and code generators, which speed up computer processing. The cloud may be the most popular option for RDBMS, but firms will consider other technologies too.
Leverage Cloud Computing
Companies will migrate many databases to the cloud, including using Database as a Service (DBaaS), a database management resource including storage and integration, a console for management, security and access, and analytics and automation. Database managers will expect DBaaS to give better service at less cost, securely, with automation and accessibility. This will lower the barrier to machine learning training and other transformative technologies. Throughout 2020 and beyond, 42 percent of firms hope to “offload administrative burdens” through using DBaaS. DBaaS migration and use success will depend just as equally on Data Quality, Data Governance and how well the cloud service capabilities purchased match the overall Data Strategy.
Most likely, leveraging cloud computing for database management will generate mixed results. Complexity querying cloud database systems, “inflexibilities in integrations, and network performance issues” means lower cloud adoption by the end of the year.
Promises of emerging cloud-based quantum computing, a cloud-based invocation of quantum emulators, simulators, or processors, will raise questions about its integration with databases on older cloud technologies. However, the DBaaS features including data virtualization, a view of data in real-time at once, will give the simplicity and flexibility some firms need.
Include Graph Databases
In 2020, database managers and administrators will strongly consider adding graph databases to their database systems portfolio. Graph databases provide quick views of data relationships that can be assessed quickly, “with less compute power overheads.”
DATAVERSITY found that 22.58 percent plan to adopt graph databases in the next year or two. But a small number of companies will adopt graph databases next year, to do complex analysis and train algorithms. Graph database will gain traction. Kurt Cagle at Forbes says graph databases will become the “go-to database of the 2020s.” They will be a necessary tool for BI in the 2020s.
Use Multi-Database Systems MDDBMS
Database managers and administrators in 2020 will handle a patchwork quilt of database systems with different schemas, including data warehouses and NoSQL. In a survey performed by Scalegrid, “44.3 percent of the responding organizations reported using multiple databases, including relational and NoSQL databases.” The number of firms using multiple databases will grow in 2020 as this becomes the norm.
Multi-database systems keep database management simple in that each database can keep its organizational schema while querying the whole group. Database servers can be added and removed through MDDBMS as needed, allowing for some flexibility. These advantages will continue to make MDDBS an attractive option for database managers and administrators in 2020.
Provide Machine Learning Infrastructure
Providing machine learning database infrastructure will be challenging for some companies in 2020. To be successful, machine learning requires higher volumes of data from multiple sources in a shorter time frame.
On top of that, data needs to be reliable and trustworthy so that the algorithms identify and use data patterns correctly. This means integrating available data from multiple systems seamlessly to feed to machine learning programs. To succeed, database management will need to do some groundwork first.
Firms that have already integrated their platforms and have other pieces in place, like in the CEMEX case study, will have an edge. Many other companies will likely take a path similar to McCormick, in 2020, where they don’t know what data is going to have to be brought in for training, and machine learning projects are frozen because of cost.
Moving databases to the cloud, to handle larger data volumes faster and take advantage of knowledge graph databases, may provide a boost in machine learning infrastructure. However, it is not surprising that Gartner predicts only:
“15 percent of use cases leveraging AI techniques, like machine learning, will be successful in 2020, as Data Quality and Data Governance issues will need to be resolved first.”
Augment Data Management
Database management will require heavy lifting in 2020, with less time for manual tasks. To adapt and keep database management simpler, tasks like Data Quality and Metadata Management will be automated with augmented Data Management. With machine learning analysis and programming, Metadata Management will power AI and machine learning algorithms. Furthermore, augmented Data Management will allow databases to be self-tuning and correcting.
Gartner predicts that through the end of 2022, Data Management manual tasks, including some database management ones, will be reduced by 45 percent through augmented data management. So, database administers will continue to use and expand on augmented data management to simplify database management.
Database management will be a mosaic of old and new technologies. Many companies will continue to embrace traditional relational database management practices. But 2020 will demand more database management flexibility and simplicity to remain competitive.
New technologies in 2020, when applied with good Data Governance and a positive data culture across the company, promise to increase database storage space, make databases more flexible, and decrease retrieval speeds. Many are moving their relational database systems to the cloud, to improve database management next year.
However, for some, this strategy will add to many disparate systems, increase database management complexity, and end up unsustainable by the end of 2020. Meanwhile, as other companies become more data-driven, the pressure to modernize and transform database management technologies will become greater.
Image used under license from Shutterstock.com