Birdstep RDM Embedded 8.0

Birdstep Technology (OSE: BIRD) released RDM Embedded (RDMe) v8.0, which is a powerful, cross-platform, small footprint database designed for resource-constrained environments with demanding performance requirements. RDM Embedded gives embedded applications extreme transaction rate and complex data modeling capabilities, solving the most complex applications without draining power or resources. The database offers a strong foundation for application development with tools for performance enhancement and database customization.

New Features

  • Hybrid In-Memory Database Support
    Adding in-memory database capabilities adds a huge amount of flexibility to an embedded db system. RDMe’s embedded database technology can now be configured to run completely diskless or in hybrid mode where the application designer puts part of their implementation in-memory and other parts on-disk. As RAM is getting cheaper in-memory databases are getting increasingly popular. Combining both traditional disk based database operations with in-memory db operations in a single system allows for extreme performance and flexibility.
  • True Varchar Support
    RDMe now adds support for an efficient implementation of storing variable sized strings. Fast embedded database engines implement fixed sized records, but with this new addition RDMe combines the fixed sized records proven over the last 20+ years of deployment with a high-performance variable sized string implementation.
  • Development Tools Additions
    With the addition of a clustering and de-fragmentation tool the RDMe toolkit just got even more impressive. RDMe offers an extensive development kit, with over 20 tools and utilities.
  • Sparse Indexing Support
    Embedded database engines are all about data indexing. For on-disk efficiency a data index is a duplicate but ordered instance of your data. Duplication of data steals both CPU and I/O cycles, so to avoid this overhead RDMe adds a user configurable sparse indexing system for strings. The sort order of most strings is usually resolved by comparing the first few bytes of the data. This feature addition allows you to accomplish just that, including only the number of bytes needed to resolve the string ordering in your index nodes. If the indexing sub-system requires more bytes, it will find them in the already referenced data nodes.

More info: Birdstep Technology, Inc