DZone Research: Hot Databases!
DZone Research: Hot Databases!
While Graph is mentioned most frequently, organizations are pursuing a polyglot, hybrid, multi-model, cloud strategy to get the best of breed for the job at hand.
Join the DZone community and get the full member experience.Join For Free
Download "Why Your MySQL Needs Redis" and discover how to extend your current MySQL or relational database to a Redis database.
To gather insights on the current and future state of the database ecosystem, we talked to IT executives from 22 companies about how their clients are using databases today and how they see use and solutions changing in the future.
We asked them, "Are you seeing increased adoption for one kind of database over another?" Here's what they told us:
- Graph used for certain use cases and aspects. It’s an important arrow in the quiver but it’s just a piece of the puzzle. Whereas Document is more of a complete solution. More flexibility and agility.
- We haven’t seen a lot of adoption of Graph. Some want support for Jazz. Have bi-modal to solve particular problems. From a real-time perspective, you tend to find you’re not doing full graph transversals. Most people are using since transactional needs are already being met. Look at how Facebook does secure analytics they scale on a single server with 2TB of memory.
- Graph from slow growth to the steeper growth of Graph databases and applications. Also, see businesses clearer about what they want to achieve and that drives how you architect. Partnering more with technical side to articulate what they want to accomplish (i.e. reduce fraud by 20% over the next five years). Becoming more business savvy articulating the requirements.
- Rising tide but growth rates are not the same. The newer stuff is growing faster than the stuff that’s been around for 30 years. Graph rocketing because of microservices.
- Adoption net new eyeballs Graph is getting interesting. Still in the trough of disillusionment but coming out of it. Practical talks on graphs. Users realizing class of problem hard to solve without Graph.
- The adoption of Document databases such as MongoDB has exploded. This is in part due to the rise of JSON APIs. JSON has become the data format of choice for developers, and document databases make it easy to store and query JSON. Developers no longer have to struggle with complex object-relational mapping. They can simply store their JSON objects directly in the database, making it easier to reason about data and accelerating the rate of development.
- Increasingly, we are seeing enterprises adopt a NoSQL strategy in their effort to modernize their data platforms. The typical choice is Document databases over other options as these tend to be easier to program to, scale elastically, are cloud and multi-cloud friendly. Mobile applications are another consideration, and databases that support today’s mobile applications naturally have the edge. Traditional databases have their place and continue to be used wherever necessary, but with more and more applications producing and consuming semi-structured data, more of the newer workloads are coming to document databases as they offer flexible schema and naturally fit application requirements. This, along with some of the deployment capabilities mentioned earlier, make the document databases a popular choice.
- SQL for operational. Graph is in between for operational or analytical. Getting more mature where they play and how they do. On analytics side GPU databases. All seeing growth. Modern data warehouse growing faster.
- SQL databases are more popular than ever before. SQL is the most comprehensive and widely used database language that can simply offer more capabilities than any other language. While SQL is back, it is also important to have Key Value functionality in a database, especially for newer apps.
- Companies are looking at using them in combination. Use as needed. Hybrid with a mix of platforms. See the most adoption of cloud-based platforms. Scale up and down as needed. Much more cost effective. JSON more prevalent in the data platforms.
- Question for the broader database market. DB Engines show time series as the fastest growing for the past two years. Providing SQL interface and full SQL is the right thing for this problem.
- We’re still in a period of time where customers are deploying best of breed especially creating microservices and choosing the database that is right for their needs. People will try to consolidate engines for the same data models but different data management tools for different uses.
- Rather than seeing a change in which type of database companies adopt, we’re seeing a shift in the number and different types of database companies adopt. There’s a move to polyglot environments, where companies have a relational database like SQL Server to handle transactional data, along with a NoSQL database to handle big data or act as a sandbox for the development of a greenfield application where the precise requirements haven’t yet been pinned down. We believe that the ACID properties of relational databases like SQL Server (Atomicity, Consistency, Isolation, Durability) will make them the preferred option for many companies, particularly where online transaction processing takes place. At the same time, companies will also use other databases like NoSQL when, for example, they want to query large datasets for Business Analytics.
- We are known for our multi-model capabilities and most of our customers use multiple models in a single datastore rather than using one or the other. So there is no “one kind of database over another,” it’s all the same. Different models are good for different things, but you often need multiple models within a single application, and you need to access the same data with multiple models...all at the same time. Across different data models, I am seeing more adoption of graph models and time series models.
- We’ve also seen accelerated adoption of open source and NoSQL databases. There’s an expectation that platforms can provide a wider range of capabilities and address a broader range of use cases and workloads that these non-relational technologies have enabled. This includes support of new data types/multiple data models, in-memory, data virtualization, support for distributed storage, and extended capabilities such as graph and spatial.
Here’s who we talked to:
Opinions expressed by DZone contributors are their own.