The resource-optimized high-performance database system Transbase® impresses with its 4S Technology and a wide range of integrated options that optimally support the special requirements of a wide range of applications and can be combined as required. Its lean and efficient implementation makes Transbase® a high-performance complete system for mobile applications, edge computing on IoT end devices or large server installations.
Scalable (1S) through its flexible Architecture and efficient Implementation
The quality of the algorithms and data structures used ensures excellent scalability - from a Mobile Device to Database Server. A Database Grid can also be set up using Database Replication.
Smart (2S) through Parallelization and optimized Algorithms
Transbase® is distinguished by its intelligent query optimization and consistently exploits the clustered data storage and indexed access paths. The possibilities of modern multicore architectures are used not only for the parallel processing of multiple queries, but also for the internal parallelization of individual queries. Global and local data caches speed up the processing at a small size.
Slim (3S) through its memory-like Data Structures
Efficient algorithms and data structures provide for a minimal amount of resources from Transbase® - both on the hard disk and in main memory: 2 MB of process size and 4 MB of cache already result in a very high-performance overall system. The logical and physical compression of the data at the side level significantly reduces the required memory space and thus also the IO times.
Secure (4S) through consistent Data and Access Protection
Both the storage of the database contents as well as the communication between database and application are encrypted. The database contents can be protected differently by means of authentication and access privileges at the table and view level.
- easy installation
One of the outstanding features of Transbase® is its ease of installation, which makes Transbase® an embedded database system along with an application without the end user of the application even noticing the database system in operation. The Transbase® setup can also be integrated into the setup of the application.
- administration- and maintenance-free operation
All administration functions, like e.g. Starting the service, applying changes, or backing up the database contents are available through a programmatic interface and can be integrated directly into applications.
All Transbase® operations are compulsorily organized as transactions and thus fulfill the database theory required ACID properties (Atomar, Consistent, Isolated and Durabel to be). Unlike many other DBMSs, this also excludes so-called DDL operations - i. Changes to the database schema - with.
- Before Image technique
This procedure writes a 'Before Image' for each changed page, which is used for recovery in the event of an error. When a transaction is committed, all its changes are persisted to disk. This method is suitable for predominantly reading operation.
- Delta Logging technique
This technique is specifically designed for large caches and competitive write operations because the cache does not have to be written persistently at each commit, but only a log with changes. At the same time, this log also serves to restore the most consistent possible database state after disk failures and enables flexible database backup processes during operation.
- Data and Access protection for the encrypted storage and transmission of data
- Developer Support with special queries and tools to help application developers
- Hybercube as a special index for data warehousing, OLAP and multi-dimensional analyzes with multidimensional and hierarchical clustering as well as special evaluation strategies for Data Warehouse applications (typical slice-and-dice operations)
- Dynamic Multithreading with parallel processing of multiple processes and optimal use of existing hardware resources
- Replication & Database Grids for load balancing on web servers and high-performance applications
- Fulltext Search for the efficient search for complex patterns in text documents
- more sophisticated SQL implementation providing strong typing, referential constraints, triggers and multiple index types
- parallel access to private (embedded) and remote (server) databases via distributed queries
- full transaction recovery and database replication features
HyPerf is a benchmark designed by Transaction Software to evaluate INSERT and SELECT performance of Transbase® on a large, multi-dimensional table. Multi-dimensional tables are very improtant for today's applications. Smartphones collect many data sets such as fitness or health data. These data have at least a geographical and a time dimension and associate measured values with each search point.
The benchmark consists of INSERT queries to see what data rates can be inserted as well as SELECT queries to see how data analysis queries perform. In addition, the benchmark provides a query MIX which concurrently generates INSERT and SELECT queries. The benchmark also generates and measures both single-threaded and multi-threaded load.
HyPerf demonstrates the outstanding Performance and Scalability of Transbase®. In particular, Transbase® shows better insertion rates than X (the other tested DBMS) both for single-threaded and multi-threaded Applications, and the numbers show that the Scalability is much better for Transbase® when data volume grows.
Multidimensional range queries also show much better results for Transbase® than for X, especially when the search is restricted to smaller and smaller sub-rectangles. This is true both for single- and multithreaded Applications.
Finally, mixed operations of INSERT and SELECT have been performed to assess the synchronization issues between reading and writing transactions. Both systems provide good concurrency, but Transbase® performs relatively better than X.