VoltDB is a developer of an in-memory database platform intended to support applications at the intersection of 5G and video stream processing. The company's platform interacts on unlimited streams of inbound data that makes per-event, data-driven decisions on live data; performs real-time analytics on fast-moving data; and exports data to longer-term analytics stores as fast as it arrives.
Graphic representation of VoltDB's architecture.
VoltDB's commercial in-memory DBMS is scalable and ACID-compliant. It is designed to be a NewSQL relational database to work with OLTP queries and a distributed database with sharing and data replication. The database is written in Java and C++, and supports both, as well as Erlang, Go, PHP, and Python. The platform is designed to support telecommunication applications for the deployment of 5G and works to provide low-latency and related capabilities for BSS/OSS, customer management, and revenue assurance applications to prevent revenue loss without compromising on data accuracy. This can be deployed in multiple styles: on-premise, cloud (both containerized and virtualized), and with Kubernetes integrations.
VoltDB works to offer features to make the data platform suited for telecommunications companies and the software vendors that serve them. This includes developing capabilities for non-negotiable, application-specific scale and latency challenges, without compromising on accuracy and reliability. Capabilities include reliable low-latency, stream integration, ACID-compliant transactions, intelligent real-time aggregations, horizontal scalability, cloud native deployments, high availability, active-active cross data center replication, disk persistence, and stack simplification.
Application-level features
ACID-compliant transactions
ACID compliance is considered a key for consistency, especially as 5G makes data integration faster, and consistency without sacrificing speed or performance becomes difficult. Ensuring ACID-compliance is an important feature for NoSQL-based data platforms and is often a requirement for enterprise applications.
Horizontal scalability
Horizontal scalability is considered an important feature by VoltDB as it allows the company to scale their in-memory data platforms both vertically (by adding CPU cores) and horizontally through partitioning to increase the capabilities of the database.
Real-time aggregations
The ability for a database to make intelligent, real-time aggregations for measuring, monitoring, and detecting deviations in important KPIs is considered an important feature for the NoSQL database, especially as decisions tend to be based on historical and near-recent records to attain an understanding of an event.
Reliable low-latency
VoltDB intends this to work with 5G, IoT, and machine-to-machine communication systems, especially around BSS and customer management, to enable applications to ingest data and make decisions on the scales of milliseconds without compromising data accuracy.
Stream Integration
Stream processing integration is intended to allow users to process, store, analyze, and act on data real time, as data flow in, and without needing to wait for it to arrive in batch form.
Operational-level features
Active-Active corss data center replication
The active-active cross data center replication provides two (or more) databases to replicate changes in real time, propagating changes between each other, to avoid issues such as conflict resolution, and offer better recovery and overall performance.
Cloud native
Being a cloud-native architecture can allow users to take advantage of the distributed, flexible nature of public clouds in order to shed layers of infrastructure (such as networks, servers, and operating systems) to accelerate the development and deployment of new services through continuous integration and deployment to scale resources up or down to respond to traffic spikes and one-time events.
Disk persistence
Ensuring disk persistence means data can survive after the process which created it has ended, and means that data is continued to be stored after any system is powered off, meaning users data will not go "dark."
High-availability
VoltDB suggests high availability, to reduce downtime and implement automatic recovery from system failures and a more robust system.
Stack simplification
With the speed and data quantity of 5G and IoT, the development of databases through layering does not address the data ingestion and analysis required, and VoltDB works to simplify the stack to manage ingest-to-action in milliseconds.
The VoltDB Partner Program works to provide organization the resources and technology needed to take advantage of industry disruptions and to help organizations address the demands of connected services. The Partner Program offers tailored engagement models, designed to work with organizations and their needs. This includes OEMs, value-added resellers, system integrators, and technology partners.
For these partners, VoltDB offers technology for database and stream processing for real-time applications as well as the development of 5G and fast data services for monetization, fraud prevention, customer value management, and digital transformation initiatives. The partner program offers support from dedicated partner management teams and success programs and works to help partners expand their customer base while being designed to augment big data investments, including NoSQL, Hadoop, and Kafka.
For Openet, a supplier of real-time business support systems to communication service providers, VoltDB was selected as a data platform capable of supporting the real-time demands of mobile data. VoltDB's solution offers Openet the chance to offer transactional, database-oriented applications against data feeds. And this was capable of transaction latencies of less than 20 milliseconds.
For Huawei, while developing the company's FusionInsight, an enterprise-class distributed Big Data solution and analytical platform, the company decided they needed to update its database capabilities to keep up with the increasing volume and velocity of transactional data. After trying different databases, but none that could support the millions of non-trivial queries per second and generate decision scores in single-digit milliseconds, they ended on VoltDB, which was capable of providing Huawei with the capabilities and low-latencies it was looking for. VoltDB also offered Huawei an opportunity to automate in real-time, for in-event fraud detection and the prevention of loss.
