Trending in Database
The databases most frequently mentioned were graph, SQL, document, time-series, and columnar.
Join the DZone community and get the full member experience.Join For Free
To learn about the current and future state of databases, we spoke with and received insights from 19 IT professionals. We asked, "Are you seeing increased adoption for one kind of database over another?" Here’s what they shared with us:
- It depends on what the customer is trying to do. Maturation of using the right tool to solve the problem. Easier to enable folks when they have a clear picture of what and why they are doing something.
- Over time, databases have evolved full circle. 30 years ago (when I started) I worked with VSAM Files and Hierarchical databases (like IMS). They were great for high volume transactions but not very flexible. These databases were usurped by relational databases which are still hugely popular today. RDBMS’s provided great flexibility (as they managed primary and foreign relationships between entities), but as the volumes of data increased, performance suffered. If you wanted to provide summary data (for warehouse and business intelligence) you had to de-normalize the data and load them into OLAP cubes and other tools. We have now returned to databases that are fit for purpose, for example, Graph, Document, Key Value, column stores, spatial; the difference is that we have far more flexibility, better infrastructure, in memory capabilities and the ability to scale horizontally and vertically, on demand.
- More “horses for courses” with a combination of column stores, columnar, adoption of cloud databases is much more prominent.
- While at Teradata we did a survey. What we saw still a majority of data processing happening with SQL databases = 57%. Unstructured databases were lagging behind with expectations and performance in higher teens, the graph has not matured to a state where it’s having an impact was 2%. Analytics and analytics processing was happening with graph databases. How to do in database graph analytics or graph representations of data on top of relational structure. Columnar is overtaking row-based stores even inside of SQL. Working on implementing graph constructs in SQL.
- Migrating to cloud more adoption of new SQL set of databases modern, agile, not lose rigors of a standard relational database.
- We definitely see more adoption for Graph/Spatial database due to the vast number of mobile/cell users. Users are very much tracked at every turn. His/her locational/movement data is marked and stored in some database, either for auditing or future (marketing) data trend analysis.
- Graph and time series.
- Most of everything we’re seeing is relational and document and most of it is going into either SQL Server or MySQL.
- The document data model is gaining traction as enterprises in every industry continue to look for modern database technology to handle the business challenges they are facing today and anticipate facing tomorrow. More than a third of MongoDB projects are now for relational database migrations as developers struggle to meet the demands of modern apps — multi-structured and rapidly changing data, massive volumes of new data streams, the shift to cloud, etc. This latter point is especially important as distributed databases are much better designed to run as cloud-native data layers on top of elastically scalable, global cloud platforms. The main trend is the adoption and preference for Database as a Service (DBaaS) solutions. Developers don’t want to be in the business of managing databases, they want to focus on building great applications. A DBaaS offering offloads all the administrative and operational mess to the experts, along with all the other benefits of the cloud.
- We are seeing more demand for time-series and columnar databases. In particular time-series databases, massive cloud vendors are moving into time-series. Also, in-memory solutions are becoming a lot more popular.
- It depends on the business problem people are trying to solve. 1) Growth in document databases for customer 360 with different kinds of customer data to one place — flexibility of the JSON data model and document database is important. 2) Time series is big driven by IoT and real-time analytics. A lot of customers are trying to modernize to be more scalable and more performance with Microservices and SQL access. 3) High-scale big SQL options for transitioning existing workloads.
- Still transactional for RDBMS. Faster adoption on the non-RDBMS. Data stores don’t seem to be top-of-mind for people. Oracle does an acceptable job for blob management. Get less visibility into what’s in the blob. That’s where document stores can add value but at a price. Need to know the question you want to answer to properly structure your data. RDBMS allows you to change the schema after the fact.
- The trend is toward more toward open source.
- There’s been more of dispersion of database choice on the macro scale. I think in the cloud space, it's a shift towards Database-as-a-Service, whether it’s cloud-vendor-native or third-party.
- Easily scalable database technologies are growing in adoption. Another trend is making existing technologies more scalable.
- Event databases are now in vogue. Event databases are used to capture software events from live sensors. For example, a healthcare company could be capturing lab testing data from a set of geographically distributed clinics. Or a retail company could be tracking inventory changes into a database that feeds into its ordering system. The increased adoption of event databases has to do with our improved ability to instrument more activities in the enterprise and the premium organizations are placing on understanding exactly what is going on in their businesses on a minute-by-minute basis. Going beyond real-time insights, even having more granular historical event data is particularly useful for analyzing and driving optimizations into the business.
Here are the contributors of insight, knowledge, and experience:
- Raghu Chakravarthi, SVP and Chief Product Officer, Actian
- Joe Moser, Head of Product, Crate.io
- Brian Rauscher, Director of Support, Cybera
- Sanjay Challa, Director of Product Management, Datical
- OJ Ngo, CTO, DH2i
- Anders Wallgren, CTO, Electric Cloud
- Johnson Noel, Senior Solutions Architect, Hazelcast
- Adam Zegelin, SVP Engineering, Instaclustr
- Daniel Raskin, CMO, Kinetica
- James Corcoran, CTO of Enterprise Solutions, Kx
- Neeraja Rentachintala, V.P. of Product Management, MapR
- Mat Keep, Senior Director of Product & Solutions, MongoDB
- Philip Rathle, V.P. of Products and Matt Casters, Chief Solution Architect, Neo4j
- Ariff Kassam, V.P Products, NuoDB
- Dhruba Borthakur, co-founder and CTO, Rockset
- Erik Gfesser, Principal Architect, SPR
- Lucas Vogel, Owner, Endpoint Systems
- Neil Barton, CTO, WhereScape
Opinions expressed by DZone contributors are their own.