Hard Real-Time Database Management: Studies, Insights, and Applications
This page serves the simple purpose of collecting and referencing research related to the design of eXtremeDB/rt. The materials focus primarily on hard real-time database management systems, an area where we believe there is a lack of recent work, as well as ongoing research into database management over flash memory, which has become increasingly relevant to us with the recent addition of native direct flash memory support in our product.
This page is part of a commercial website, and the selection of materials referenced here reflects what our team has chosen to include on this page. However, if you believe that something on this page should be corrected, or if you have additional material to contribute, please email us and we will review your suggestion.
On this page
Important definitions
Videos
White papers
Articles
- The Only COTS Hard Real-Time DBMS Slashes Memory Footprint and Advances flash Simulation Capabilities
- When a Millisecond Is a Promise
- Winners of the embedded award 2026
- Embedded World 2026 Winners Showcase Innovation Across AI, Security, Real-Time Systems
- 3 Reasons eXtremeDB Is the Best Database for Your STM32 System
- Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 4
- Update: eXtremeDB/rt 2.0, All the reasons why embedded systems need a database, now with persistence
- Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 3
- Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 2
- Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 1
- Real-Time Database Management on Raw Flash Memory: From Complex Infrastructure to Streamlined Drivers.
- eXtremeDB: All the reasons why embedded systems need a database and how to go about it.
- Real-Time Is a Misnomer
- Real-Time Versus Real Fast
- On Predictable Real-Time Data Management Over Persistent Storage Media. Q&A with Andrei Gorine and Steve Graves
- 11 Myths About Real-Time Database Systems
- Buy vs. Build for Time-Critical Sensor Data Fusion
- On Real-time Databases. Q&A with Andrei Gorine and Steve Graves
- The Data in Hard Real-time SCADA Systems Lets Companies Do More with Less
- Latest embedded DBMS supports asymmetric multiprocessing systems
- Real-time database systems are not actually real-time. Unless they are.
- Dogmatic views on safe languages used for mission- and safety-critical applications
- Soft v Hard Real-Time Systems
- What is and what isn’t a hard real-time database system?
- Embedded Databases in Real-Time Systems
- Keeping A Mission-Critical Deterministic DBMS On Time
- Real-time tasks need real-time data
Important Definitions
Real-time database
“Real-time database” has come to have two meanings. The first and, so far, the predominant use of the term refers to a database system which uses streaming technologies to handle workloads for which the state changes in real-time as a reflection of real-world events. This differs from traditional databases containing persistent data, mostly unaffected by time. In this context, real-time processing means that a transaction is processed fast enough for the result to come back and be acted on right away. Such real-time databases are useful for fraud detection, recommendation engines, audience targeting, and pricing strategies.
The other use of the term “real-time database” embodies a more strict definition of real-time. A hard real-time database system is one that is time cognizant: It is aware of processing deadlines and offers transaction schedulers that enforce those deadlines and are adaptable to varying workloads and changing deadlines and task priorities (e.g., works equally well with periodic and aperiodic tasks). This is the only type of database management system that, when used with a RTOS, is acceptable for use in mission-critical systems.
Deterministic database
“Deterministic database management system” has come to have two meanings.
A deterministic real-time database system guarantees that transactions complete (commit or rollback) within a set deadline. Transactions that are destined to be “late” are identified, interrupted and forced to initiate rollback in time to satisfy set deadlines. The key is to ensure that all database runtime internals are in a “recoverable” condition.
In the context of distributed systems, deterministic can mean that changes are organized up front, and then applied. For example, geographically distributed replicas (or shards) can be given the same input and be guaranteed to produce the same result, making data replication (and replay) easier to implement.
Videos
Hard Real-Time Data Management Over Persistent Storage Media
This Webinar delves into deadline management for real-time data management deployed on NAND flash storage. After a brief overview of database transaction deadline management, it will examine the core issues inherent to flash media that complicate guaranteeing deadlines.
Is Hard Real-time Possible with a DBMS?
Is Hard Real-Time Possible with a DBMS? Presented by McObject CEO and Co-founder, Steve Graves. A concise, 30-minute overview of the critical factors that are required in a DBMS to succeed in hard real-time systems.
Database System Requirements to Achieve Level 3+ Autonomous Driving
Presented by McObject CEO Steve Graves at the EETimes Advanced Automotive Tech Forum.
Real-time is not the same as real fast. Speed is not the most important metric for real-time systems like ADAS – deterministic response time is. Speedy response is desirable, but speed alone does not guarantee timely response from a database system. For the same reason that real-time operating systems are used for ADAS, a real-time database system is needed. This presentation will explain the difference between a real-time database system and conventional database systems.
White Papers
Real-time Database Management on NAND Flash Storage
In this paper, we delve into deadline management for real-time embedded databases deployed on NAND flash storage. After a brief overview of database deadline management, we examine the core issues inherent to flash media that complicate guaranteeing transaction deadlines. These challenges include wear leveling, garbage collection, and bad block management, typically handled by flash translation layers (FTLs) provided by manufacturers or independent software vendors. We then propose a practical solution for real-time database management that integrates FTL with database transaction managers, enabling effective deadline management while accessing storage devices.
Predictable Real-time Data Management Over Persistent Storage Media
Embedded designs with predictable performance requirements are subject to timing constraints – their response to events must fit within timeframes that are known in advance. The growing complexity of real-time systems translates to more, and more complex, data to manage. Further, the volume of data may be such – or the regulatory need for data retention may mandate – that persistent media be introduced. However, hard-disks, NAND flash, and SSD introduce unpredictable latency. How can timing constraints be met in this case?
Shared Data in Asymmetric Multiprocessing (AMP) Configurations
Heterogeneous multi-core systems are becoming ever more popular for automotive and industrial applications due to their high performance and energy efficiency. This paper will discuss the design and challenges of a shared data implementation for AMP configurations, explore implementation options to synchronize access to shared data, and discuss example use cases.
Real-time Deterministic Database Management
This paper will discuss the objectives of deterministic, predictable database management in the context of real-time application design. We will introduce extensions to a conventional database management systems (DBMS) transaction scheduler that add semantics, enforce database transaction priorities and deadline scheduling. We will then focus on the practical aspects of the design and demonstrate its use in several real-life application patterns in a variety of real-time operating system (RTOS) environments.
Articles
The Only COTS Hard Real-Time DBMS Slashes Memory Footprint and Advances flash Simulation Capabilities
Read the article in StorageNewsletter.com
When a Millisecond Is a Promise: Andrei Gorine on Hard Real-Time Data Management and the Hidden Cost of Flash
Read the article by Roberto Zicari in ODBMS.org
Winners of the embedded award 2026
A press release from the embedded world Exhibition & Conference committee. A high-calibre jury evaluated more than 110 entries for the 22nd embedded award. We present the winners in the nine categories of the embedded award embedded-world.de
Embedded World 2026 Winners Showcase Innovation Across AI, Security, Real-Time Systems
An Embedded World jury chose these products as the most innovative breakthroughs in embedded systems, edge computing, and automotive tech. Read the article in Design News .
3 Reasons eXtremeDB Is the Best Database for Your STM32 System
January 27, 2026. When you’re building on an STM32, you’re choosing components that must coexist with tight memory budgets, real time deadlines, and deterministic behavior. That’s exactly where eXtremeDB stands apart. Review the three reasons it’s the smartest database you can embed in an STM32 based system on LinkedIn .
Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 4
December 12, 2025. This next section turns to a new question: how real-time guarantees can be extended all the way down to persistent storage — specifically, flash memory. Embedded Computing Design .
Update: eXtremeDB/rt 2.0, All the reasons why embedded systems need a database, now with persistence
Updated December 4, 2025. eXtremeDB/rt 2.0 supports persistent databases thanks to two new middleware: xFile and xFlash. Read more on the ST Blog .
Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 3
December 3 2025. Once timing and verification are in place, the focus shifts to what happens after an enforced abort and how the database kernel keeps logical and temporal consistency intact. Read more on Embedded Computing Design .
Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 2
November 25, 2025. Part 2 reviews how the database kernel not only follows the schedule but actually meets every deadline, every time. Read more on Embedded Computing Design .
Engineering Real-Time: Lessons Learned While Chasing Determinism, Part 1
November 17, 2025. In hard real-time systems, deadlines aren’t suggestions, they draw the line between working and failing. Every task must finish exactly when it should; even a tiny slip can ripple through and throw everything else off balance. Read more on Embedded Computing Design .
Real-Time Database Management on Raw Flash Memory: From Complex Infrastructure to Streamlined Drivers.
October 17, 2025. Embedded systems are data-driven and demand structured persistence with predictable performance. Raw flash provides deterministic, low-latency access, while a real-time database management system (DBMS) supplies the infrastructure for transactions and recovery. This article demonstrates that only their combination delivers efficient, reliable storage for hard real-time tasks. Read the technical article in Embedded.com
eXtremeDB: All the reasons why embedded systems need a database and how to go about it.
July 9, 2025. What if a database for embedded systems, such as eXtremeDB, were the only thing standing between critical sensor data and a massive catastrophe? Read the Partner Program blog post on ST.com
Real-Time Is a Misnomer
April 23, 2025. In programming jargon, the term “real-time” is not used in its literal sense, which can be confusing. The literal definition means that when something happens in real time, it happens instantaneously. But in practice, something that occurs very quickly is mistitled as being real-time. Why does the misconception continue to occur, and what problems does it cause? Listen to the Embedded Executive podcast.
Real-Time Versus Real Fast
December 18, 2024. An overview of the two definitions for the term ‘real-time’ database and why we should use the original meaning. Read the complete article in Embedded Computing Design .
On Predictable Real-Time Data Management Over Persistent Storage Media. Q&A with Andrei Gorine and Steve Graves
October 18, 2024. Q1) As real-time systems grow in complexity, they require managing increasingly large and intricate data. What are the main challenges in managing the predictability aspects of real-time databases? Read the answer and complete article in ODBMS.org
11 Myths About Real-Time Database Systems
August 27, 2024. What this article covers:
– The actual definition of real-time database versus the abused interpretation used by marketers of real fast database systems.
– The interplay of transient and persistent memory, and hard and soft real-time capabilities.
Read the article in ElectronicDesign.com .
Buy vs. Build for Time-Critical Sensor Data Fusion
May 15, 2024. What exactly is “real-time”? This article looks at its permutations and offers insight into time-critical sensor data fusion and its role in database systems. Read the article in ElectronicDesign.com
On Real-time Databases. Q&A with Andrei Gorine and Steve Graves
October 9, 2023. “Another rapidly evolving concept is real-time mixed criticality systems that integrate components with different levels of criticality onto a single computing platform.” ODBMS.org.
The Data in Hard Real-time SCADA Systems Lets Companies Do More with Less
August 11, 2023. The latest database technology allows the capture and utilization of previously out of reach data found in mission-critical SCADA systems. Read the article in Automation.com
Latest embedded DBMS supports asymmetric multiprocessing systems
May 24, 2023. McObject’ latest eXtremeDB/rt 1.2 release supports heterogenous architectures with different CPU architectures executing different operating systems and have a large amount of shared memory. Read the Embedded.com article.
Real-time database systems are not actually real-time. Unless they are.
January 31, 2023. A blog post by McObject CEO Steve Graves. Many real-time systems need a database. However, choosing one can be challenging depending on performance requirements. Find the full text at ElectronicDesign.com
Dogmatic views on safe languages used for mission- and safety-critical applications
December 20, 2022. A blog post by McObject CTO Andrei Gorine, highlighting a few challenges and sharing a few lessons learned about memory management in embedded applications for safety-critical system development, reliability and dependability. Find the full text at Embedded.com.
Soft v Hard Real-time Systems
November 4, 2022. This short article emphasizing distinctions important to developers of safety- and mission-critical hard real-time systems. Read more at EmbeddedComputing.com
What is and what isn’t a hard real-time database system?
May 29, 2022 blog post. Surprisingly, it’s not as simple as it sounds, but perhaps we could start by defining a few terms. Learn more at DB-Engines.com
Embedded Databases in Real-Time Systems
March 14, 2022, Real-time systems are evolving at an incredible rate. Real-time systems used to be relatively simple, like the anti-lock braking systems in aircraft and, later, automobiles. Today, real-time systems are more complex. Advanced Driver Assistance Systems (ADAS) are kind of the poster child. Learn more in the article in Circuit Cellar.
Keeping A Mission-Critical Deterministic DBMS On Time
September 20, 2021 ,Open-source and commercial database management systems (DBMSs) have been used in embedded systems for more than 20 years. For the most part, developers don’t even debate whether to buy or build their own DBMS for embedded systems; that component is almost always bought. Get the news from EmbeddedComputing.com
Real-time tasks need real-time data
Hard real-time systems are following the same trajectory as embedded and soft real-time systems: They’re more elaborate systems that are being required to do more with more data that’s more complex. Historically, real-time system developers have had to separate database management from hard real-time tasks, do without, or develop their own real-time database management. Read the Wind River blog.
Academic studies and external papers
Optimizing Robotic Precision Unleashing Real-Time Performance With
Advanced Foundational Software Solutions
March 2025. This QNX white paper explores the challenges of latency and jitter in robotic systems and the importance of deterministic behavior in safety- and mission-critical applications. It describes how real-time operating systems use predictable scheduling, deterministic interrupt handling, and microkernel-based architectures to improve the precision, reliability, and responsiveness of robotic control systems.arXiv.org
How to Write to SSDs
March 10, 2026. This paper demonstrates that adopting out-of-place writes is essential for database systems to fully leverage SSD performance and extend SSD lifespan. We propose a set of out-of-place optimizations that collectively reduce write amplification across both the DBMS and SSD layers. We redesign the in-place, B-tree-based LeanStore to write out-of-place and support these optimizations, and evaluate it on diverse OLTP benchmarks, dataset sizes, and SSDs. Review the paper at arXiv.org
Towards Optimal Transaction Scheduling
July 2024. Maximizing transaction throughput is key to high-performance database systems, which focus on minimizing data access conflicts to improve performance.However, finding efficient schedules that reduce conflicts remains an open problem.This work proposes systematically exploring the entire schedule space, proactively identifying efficient schedules, and executing them precisely during execution to improve throughput. VLDB.org
Schedulability Analysis in Fixed-Priority Real-Time Multicore Systems with Contention
9 May 2024. In the scheduling of hard real-time systems on multicore platforms, significant unpredictability arises from interference caused by shared hardware resources. The objective of this paper is to offer a schedulability analysis for such systems by assuming a general model that introduces interference as a time parameter for each task. Read the paper on MDPI.com
An experimental comparison of different real-time schedulers
1 October 2012. In this work, an experimental comparison among the Rate Monotonic (RM) and Earliest Deadline First (EDF) multiprocessor real-time schedulers is performed, with a focus on soft real-time systems. We generated random workloads of synthetic periodic task sets and executed them on a big multi-core machine, using Linux as Operating System, gathering an extensive amount of data related to their exhibited performance under various real-time scheduling strategies. The comparison is based on the fixed-priority scheduler for multiprocessors as available in the Linux kernel, and on our own implementation of EDF, both configured in global, partitioned and clustered mode. The impact of the various scheduling strategies on the performance of the applications, as well as the generated scheduling overheads, are compared presenting an extensive set of experimental results. These provide a comprehensive view of the performance achievable by the different schedulers under various workload conditions. Read more in Science Direct .