메인으로

Altibase CTO: "We Will Sequentially Release Altibase Windows 2026, Sharding 3, and 8.1 by Q1 Next Year"

"We plan to sequentially release Altibase Windows 2026, Altibase Sharding 3, and Altibase 8.1 by the first quarter of next year."

Lee Seung-won, Chief Technology Officer (CTO) of Altibase's Research and Development Division, recently told ZDNet Korea in an interview, "We are driving product innovation centered on AI, cloud, and distributed processing technologies."

Altibase is a domestic pioneer in hybrid DBMS. Hybrid DBMS refers to an architecture that uses both in-memory tables (residing in memory for fast processing) and disk-based (on-disk) tables within a single interface.

CTO Lee Seung-won emphasized, "Altibase is the company that released the world's first commercial hybrid DBMS integrating in-memory and on-disk DBMS in a single engine. It enables ultra-high-speed performance and large-capacity data processing, with further enhanced performance in cloud environments." He continued, "According to our own benchmark test (BMT) results, we recorded performance up to 4-5 times faster than competing legacy DB products. Particularly in large-volume transaction processing and complex query environments, we maintained consistent high performance, recording clear advantages over competitors in both data loading and query speeds. This achievement results from the combination of in-memory-based architecture and optimized hybrid DBMS structure. We anticipate significant performance improvements in environments requiring large-scale real-time data processing, such as finance and telecommunications."

Recently, security vulnerabilities have occurred successively in domestic IT companies and some DBMS products, raising concerns across the industry. In this situation, Altibase is focusing on securing product stability and reliability through preemptive security enhancements and advanced management systems. CTO Lee stated, "In particular, through the 'Altibase Backup Manager' scheduled for release next year, we plan to provide backup file encryption functionality and support security features that prevent direct password exposure by utilizing encrypted password files."

Previously, Altibase drew attention in July by announcing a three-year phased technology roadmap. This roadmap, running until 2027, consists of four major areas: implementation of in-memory large-capacity high-performance technology, cloud-native database, AI agent and vector database functionality, and distributed cluster technology. Below is the Q&A interview with Altibase CTO Lee Seung-won.

What Kind of Company is Altibase?

"We are the company that developed the world's first hybrid DBMS. It enables ultra-high-speed and large-capacity data processing in databases. The latest version is 'ALTIBASE V7.3,' with further enhanced performance in cloud environments. In particular, 'ALTIBASE V7.3' has been successively supplied to mission-critical systems of global financial and telecommunications companies, proving the global competitiveness of Korean DBMS products."

What are the Main Features and Advantages of Your Core Product, Hybrid DBMS?

"There are several competitive advantages. First, the world's first Hybrid Partition functionality. Altibase 7.3's hybrid partition technology has a unique architecture that integrates in-memory DB and disk-based DB engines into one platform. It can automatically or selectively operate memory partitions and disk partitions within a single table according to data characteristics, efficiently managing the ultra-high-speed processing performance of in-memory DB and the large-capacity storage capability of disk DB at a single logical table level.

This technology satisfies the conflicting requirements of real-time transaction processing and stable large-capacity data storage simultaneously, and is evaluated as a rare innovation case globally in terms of cost-efficiency and technical uniqueness.

Second, dramatic performance improvement. 'Altibase 7.3' achieved performance improvements of up to 490% or more compared to the previous version. This results not simply from hardware performance dependency, but from fundamental improvements to bottlenecks in commit algorithms and logging processing, which are core to transaction processing. This performance innovation stably supports the maximized throughput required in industrial sites such as telecommunications and finance where ultra-high-speed transaction processing is essential.

Third, cloud environment optimization and Kubernetes-based scalability. 'Altibase 7.3' provides AKU (Altibase Kubernetes Utilities), a cloud-specific utility, to automatically perform auto scale-out and scale-in. This enables elastic response to sudden traffic fluctuations, and parallel replication functionality during node increases has also significantly improved expansion speed.

Fourth, global redundancy and DDL (Data Definition Language) synchronization technology. By applying high-speed lossless compression technology to replicated data, global redundancy performance has been greatly improved. This enables stable real-time transmission of large-capacity data even in long-distance WAN environments such as Los Angeles to New York (approximately 4,500 km), providing capabilities optimized for building distributed systems for global companies.

In particular, Replication DDL synchronization technology is Altibase's proprietary function that guarantees atomicity of DDL operations across all nodes. It fundamentally prevents master-slave structure inconsistency problems commonly occurring in existing DBMS, allowing administrators to safely perform schema changes without separate complex measures.

Fifth, enhanced compatibility and user convenience. We focused on user-friendly design. We provide a dedicated migration center tool for PostgreSQL to easily support migration from existing PostgreSQL environments to Altibase, and also provide a plugin for DBeaver, a universal DB management tool, enabling various tasks such as database construction, management, and query writing in an intuitive graphic environment."

Who are Your Major Customers?

"We have many customers in various industries including public sector, finance, telecommunications, and manufacturing. Representative customers in the public sector include the Ministry of the Interior and Safety, Ministry of Economy and Finance, Ministry of Oceans and Fisheries, Ministry of Land, Infrastructure and Transport, Ministry of Trade, Industry and Energy, Ministry of SMEs and Startups, Ministry of Public Safety and Security, Ministry of Employment and Labor, Ministry of Food and Drug Safety, Fair Trade Commission, Ministry of Personnel Management, Korea Social Security Information Service, Korea Airports Corporation, Korea Rail Network Authority, Korea Workers' Compensation & Welfare Service, Korea Transportation Safety Authority, Incheon International Airport Corporation, LH, KORAIL, Korea Petroleum Pipeline Corporation, Seoul Metro, etc.

In finance, our products are used by Woori Bank, Mirae Asset Securities, Shinhan Bank, Korea Exchange, Kakao Pay Securities, Samsung Fire & Marine Insurance, Korea Investment & Securities, Eugene Investment & Securities, Yuanta Securities Korea, Daishin Securities, Kookmin Bank, Kiwoom Securities, Shinyoung Securities, KB Securities, NH Investment & Securities, Dongbu Securities, Kyobo Securities, Hana Capital, Bookook Securities, etc.

In telecommunications, all three major carriers - SKT, KT, and LG U+ - are our customers. In manufacturing, customers include Samsung Display, S-Oil, LS Cable & System, HP, LG Electronics, Dongwon F&B, Samsung SDS, Samsung SDI, SK Siltron, CJ, SK hynix, etc. International customers include AT&T, Telus, MobiCom, HP, Digi, TELUS, NOMURA, InnoCaption, etc."

V7.3's Slogan is 'Mission-Critical Succeed.' What Does It Mean?

"It means we will successfully complete (Succeed) customers' core mission-critical tasks using Altibase technology. It's a slogan emphasizing that Altibase's in-memory DBMS technology greatly contributes to stable operation and fast processing of critical systems that must never stop, such as financial payments and aviation control.

This slogan encompasses three core values. First, that companies' critical tasks must be completed successfully without interruption; second, that Altibase's in-memory DBMS technology provides very fast data processing speeds to effectively support these core tasks; and finally, our customer-oriented promise to lead enterprise customers to secure competitive advantages and succeed in business by processing large volumes of data quickly and efficiently through Altibase technology in the rapidly changing IT environment."

You Announced a Three-Year (2025-2027) Technology Roadmap in Late July. What Does It Include?

"It contains our commitment to implement phased advanced technologies over the three years from this year to 2027. It's also a declaration of our full-scale entry into the global market. We established it reflecting rapidly changing domestic and international market requirements and AI technology development trends. It consists of four core technologies: in-memory large-capacity high-performance technology, cloud-native database, AI agent and vector database functionality, and distributed cluster technology.

First, we plan to sequentially introduce Altibase Windows 2026, Altibase Sharding 3, and Altibase 8.1 by the first quarter of next year. Simultaneously, we are strengthening R&D in AI, cloud, and distributed technology fields to enhance product competitiveness. In particular, in the second half of 2027, we plan to launch next-generation products with full-scale application of real-time distributed cluster technology, providing customers with even faster performance and scalability."

Vector Databases are a Hot Topic in AI Recently. What Preparations is Altibase Making?

"When processing large-scale data in existing AI services, data was distributed and stored using NoSQL distributed memory DBs like Redis. However, specialized vector databases supporting indexing and vector operations are emerging and attracting attention. Since there are no standards yet, development interfaces differ for each product and usage is complex, making it difficult for developers to utilize. Altibase is researching technology to support vector operations and indexing based on existing SQL to solve these problems.

In next-generation Altibase, we plan to enable vector operations through standard SQL syntax and familiar interfaces like ODBC (Open Database Connectivity) and JDBC (Java Database Connectivity), and to enable developers to easily utilize AI and vector search functions without separate learning."

Please Explain 'Altibase Sharding 3' Scheduled for Release Next Year

"Data is rapidly increasing in the AI, big data, and IoT era. Especially, real-time processing is often needed for sensor data. Previously, real-time capability was secured by distributing data with NoSQL memory DBs like Redis. However, this has limitations in complex computation processing and consistency management.

Altibase provides faster performance and SQL compatibility than NoSQL as a memory DBMS. However, there was a constraint of difficulty containing large-scale data due to memory capacity limitations. Customers made many requests to 'use Altibase in distributed environments.' Altibase already had distributed sharding functionality, but there were cases where update operations were interrupted during node expansion. This Altibase Sharding 3 resolved these problems. Online node expansion through distributed sharding is possible. In particular, 'Altibase Sharding 3' launching in January next year can freely expand nodes even as data grows, providing both stability and performance in large-scale service environments."

How is AS3 (Altibase Sharding 3) Specifically Different from Existing Altibase Products?

"Existing Altibase primarily had a structure that enhanced performance through vertical scaling. However, as data and traffic increased exponentially, the need for horizontal scaling grew. Existing Altibase sharding had a 'meta node' special node, and all connections had to access this meta node, creating a problem of concentrated load.

'AS3' resolved this problem by adopting a structure where all nodes perform identical roles (No special node). It supports shard key-based data distribution and online resharding. Thanks to this, performance can be scaled almost linearly by simply adding nodes. Problems of data concentration on specific nodes can also be resolved. Even during failures, other nodes immediately take over roles, enabling stable service provision without SPOF (Single Point of Failure)."

Why is Altibase Launching Clusters Including Sharding Functionality Through 'AS3'?

"Sharding is core technology for efficiently managing rapidly increasing data and traffic. Distributing data across multiple nodes prevents performance degradation and bottlenecks, and enables linear scaling by adding servers when needed. Simultaneously, specific node failures don't lead to complete service interruption, strengthening business continuity.

From a cost perspective, 'sharding' also has significant advantages. Servers can be added at lower cost than disk-shared clusters, expecting TCO reduction effects. Actually, major domestic large enterprises have consistently requested sharding for large-volume transaction processing and real-time analysis. Overseas, partners in Middle East and Southeast Asian regions requested key-based data distribution structures similar to MongoDB and Redis.

'AS3' is a solution developed reflecting these domestic and international demands. We hope it will establish itself as essential data infrastructure in IoT, big data, and AI environments."

What are the Specific Characteristics of AS3's Sharding Method?

"AS3's biggest differentiator is that all nodes perform equal functions. Even without separate coordinator nodes, SPOF (Single Point of Failure) doesn't occur, and even during failures, service continuity is guaranteed through automatic reconnection functionality.

It's also powerful in terms of scalability. Capacity increases linearly with each node addition, which was proven in actual tests. Data distribution supports hash and range methods based on shard keys, and data can be redistributed without interruption during operation through online resharding functionality.

Developer-friendliness is also an advantage. Since it supports existing DBMS interfaces including SQL, ODBC (Open Database Connectivity), and JDBC (Java Database Connectivity), existing applications can be migrated with minimal changes. Companies can minimize operational risks while adopting new technology."

What Should Companies Adopting 'AS3' Prepare?

"First, compatibility review with existing systems is necessary. Data migration planning and operational infrastructure inspection are also required. Particularly in sharding structures, performance varies greatly depending on which value is designated as the shard key, so understanding of data modeling and distributed query optimization is essential. Altibase plans to provide end-to-end services from preliminary consulting to migration support, training programs, and operation guides so customers don't experience difficulties in these areas. Through this, customers can reduce initial trial and error and quickly establish stable operational systems."

Cloud Strategy... Native JSON Support in Next-Generation Altibase 8.1

What Efforts are You Making to Provide DBMS Optimized for Cloud Environments Regarding Altibase's Cloud Response?

"The cloud core is 'flexibility to pay only for what you use.' To this end, Altibase supports easy node expansion and reduction in cloud environments through 'Altibase Kubernetes Utility (AKU).' When usage increases, nodes expand and data is automatically replicated. And when load decreases, nodes are removed again, providing operational efficiency.

Also, the disk-sharing method commonly used by existing cloud RDBMSs distributes read operations but has limitations in distributing update operations. We improved this by developing re-sharding technology enabling online data redistribution without service interruption, and implemented it in the AS3 version launching in 2026.

Finally, cloud and big data environments have characteristics where structured and unstructured data occur together and schema changes are frequent. Reflecting this, next-generation Altibase 8.1 natively supports JSON. JSON (JavaScript Object Notation) is a lightweight data exchange format used when storing or exchanging data.

In particular, we implemented self-developed binary JSON to reduce storage capacity and greatly improved in-document search processing performance and document validation speed. Through this, faster and more stable data processing is possible in various cloud applications."

You Recently Developed 'Kafka Connector.' How is This Related to Altibase's Technology Roadmap?

"We are currently enhancing products through our three-year technology roadmap (2025-2027), presenting four pillars: cloud-native DBMS, AI and vector DB, distributed cluster, and in-memory high-performance technology. This Kafka Connector is highly significant in terms of strengthening connectivity with cloud and big data environments. In cloud, IoT, and AI service environments, real-time data streaming is core, and Kafka has established itself as the de facto industry standard. Altibase's official support for Kafka connectivity means Altibase's utilization scope will widen further within the cloud, AI, and big data ecosystem."

'Altibase Windows 2026 (AW 2026)' Attracting Attention

What Background Led to Planning 'Altibase Windows 2026 (AW 2026)' for Release at Year-End?

"In embedded environments such as manufacturing equipment, medical devices, and military/railway systems, servers and clients often need to operate simultaneously within one device. For example, MRI equipment and POS terminals must process and store data internally without external server connection, and synchronize with central servers when necessary. In these structures, the operating system (OS) is often Windows. We developed it because there were many requests to support running the server itself directly in Windows environments."

What Advantages are There in Using Altibase in Windows Environments?

"Having servers and clients together in one device enables building terminal-integrated solutions. That is, data storage/processing and user applications occur within one system, speeding response times, and data can be stably processed and stored locally even during network delays or failures. For example, in medical equipment, patient data can be immediately analyzed and stored, and in manufacturing site machinery systems, equipment sensor data can be collected and analyzed in real-time to immediately control equipment status."

Please Introduce the Supported OS and Technical Features

"'AW 2026' officially supports Windows Server 2022, Windows Server 2025, and Windows 11. It provides the existing Altibase strength of memory-disk hybrid structure as-is, enabling stable operation even on embedded devices with limited memory resources. It also supports redundancy and data replication between hundreds to thousands of devices. Through this, even when multiple embedded devices aren't always connected to servers, they can process data locally and automatically synchronize when connected to servers."

What Significance Does the Release of 'Altibase Windows 2026' Have?

"We have long built competitiveness based on performance and hybrid DBMS characteristics in Unix and Linux-based markets. However, it couldn't be used on Windows servers until now. With this release, we officially support Windows Server, allowing customers to adopt Altibase without changing their OS and experience the innovation of high-performance hybrid DBMS. This means Altibase can be utilized in more industrial sites, and is particularly significant in providing a new option for Windows environment industries performing mission-critical tasks."

Global Market Expansion... Corporations in Singapore, etc.

What Strategy are You Pursuing for Entering Southeast Asian Markets Regarding Altibase's Global Market Entry?

"In January this year, we established 'ALTIBASE SEA' corporation in Singapore, launching full-scale Southeast Asian market entry. Currently focusing on Indonesia and Thailand for Southeast Asian market entry, we are pursuing proof of concept (PoC) projects targeting local governments and public institutions, and completed our first business order from a public institution in Thailand. We are also actively exploring cooperation plans with core solution companies for ERP, payment systems, and school management systems, expecting full-scale business expansion in local markets."

What Collaborative Relationships are You Building with Overseas Companies?

"We are conducting 'In-Memory Database Appliance Development Joint Projects' with global IT companies headquartered in the Asia-Pacific (APAC) region. Recently, we completed porting to enable ALTIBASE 7.3 operation in IBM's mainframe-based LinuxONE environment. While expanding cooperation systems with overseas ISVs (Independent Software Vendors), SIs (system integration), and consulting companies, Altibase has been recognized as IBM's official ISV and is expanding global partnerships by participating in IBM partner programs held in Indonesia, Philippines, Singapore, and Hong Kong."

Please Share Global Customers and Success Cases

"Altibase has been officially recognized for stability and reliability by meeting IBM's strict standards, and this will become the foundation for greatly expanding the approximately 700 customers and 8,000 references already secured."

What is Your Differentiation Strategy for Securing Competitiveness in Global Markets?

"Based on strengths in domestic public and enterprise markets, we will present a blueprint for global market expansion and technological innovation in the AI era. Particularly in an era where AI becomes core to corporate competitiveness, Altibase plans to support companies' AI transformation as data infrastructure guaranteeing ultra-high speed, scalability, and stability. We will also meet customers' diverse needs in line with latest trends including fully managed cloud, NoSQL, and AI technology, opening new possibilities for database technology."

Latest Trends and Market Position

What is Altibase's Current Position and Strengths in the Domestic DBMS Market?

"Altibase is establishing itself as a leading domestic in-memory database management system (DBMS) company. We successfully applied our technology to a major domestic bank's AI voice phishing real-time monitoring system, contributing to financial crime prevention, which is representative achievement effectively utilizing actual AI applications. We have also proven the technical excellence of domestic DBMS by successfully operating major public institution systems including Seoul Metro, Fair Trade Commission, and Daegu Government Integrated Computing Center."

What Technical Vision Do You Have for the AI Era?

"We plan to unveil a technical vision targeting data infrastructure innovation for the AI era. Based on long DBMS development experience and data solution service know-how, the importance of data is being highlighted according to AI and semiconductor technology development. We plan to present a data-centric AI vision reflecting these trends.

We are researching to support vector data types and real-time distributed clusters for the future AI era, targeting fast search performance. We also plan to advance into high-performance multi-model integrated DBMS accommodating external data connectivity and diverse customer needs. We will provide dedicated connectors for vector database connectivity for unstructured data, and also provide private/public model connectivity technology through database AI agents."

What Future Major Technology Trends is Altibase Focusing On?

"Looking ahead, innovative technologies centered on AI such as agentic AI, enterprise expansion of generative AI, and expanded utilization of inference AI models are expected to lead the IT industry. Altibase is pursuing strategies to provide database environments optimized for AI applications in line with these AI technology trends. Beyond that, we are particularly focusing and accelerating strategic technology development aligned with large language models (LLM) in the generative AI era."

You Plan to Release 'Real-time Distributed Cluster Advanced' Version in 2027. What Competitive Advantages Do You Seek Through This?

"The 'Real-time Distributed Cluster Advanced' version scheduled for release in the second half of 2027 can process in-memory large-capacity data at ultra-high speed. The key differentiation point is particularly that it supports transactions impossible in existing NoSQL and document DBs, and can process even 'complex SQL statements' at high speed. Through this, Altibase has a strategy to directly challenge Redis and MongoDB, which currently dominate global markets."

What are Your Plans for Establishing Altibase as a Trusted Database Solution Provider in Global Markets?

"We will establish ourselves as a trusted database solution provider in global markets through ultra-high-performance in-memory computing-based hybrid DBMS technology. We will present a strategic blueprint to further strengthen technological competitiveness in the rapidly changing database market and solidify our position in global markets, demonstrating differentiated competitiveness in overseas markets as well."

How are You Responding to Rapid Database Market Changes?

"Through the latest version 'ALTIBASE V7.3,' we further strengthened performance in cloud environments, and are researching to support vector data types and real-time distributed clusters for the future AI era, targeting fast search performance. We also plan to advance and develop into high-performance multi-model integrated DBMS accommodating external data connectivity and diverse customer needs."

Reported by Eunju Bang | ejbang@zdnet.co.kr