Extreme scalability
Dataset size
Ancelus scales vertically to 40 Terabytes, 100 billion records, with no loss of speed. Tens of millions of transactions per second. Hardware capacity expanding 10x every 4 years. Expect 9 exabytes within a decade.
Relational systems have hit a brick wall on performance. After three decades the demands of the big data projects are beyond the reach of any relational technology without complex and temporary patches.
Hardware
Ancelus shrinks the hardware required for the most demanding projects. Typically by a factor of 4 to 10. Energy consumption is reduced by an equal amount.
Multi-core systems provide almost no relief to relational databases. With Ancelus, multi-core systems scale with near linearity
Ancelus delivers speed through the database structure, not through more server racks. Higher speeds offer opportunities for improvements in user response, concurrent users, and reduced administration downtime.
Dataspace
Ancelus reduces the dataspace size by up to 80% through two techniques:
- Elimination of duplicate data in foreign keys and concatenated keys
- Elimination of the sparse matrix problem: variable width fields are specified at the transaction level so no reserved space is lost
Locking
The combination of column level locking, spinlocks and blinding speed minimize the chance of collisions by several orders of magnitude. Column level locking allows concurrent write access to separate columns in the same table.
Spinlocks allow a 20 nanosecond lock/unlock time, and data access in under 100 nanoseconds assures that locked data is returned to service in the shortest possible time.
Shorter lock times mean more concurrent throughput and more supported users.
Did you know...
AncelusDB replaces the fundamental structure and operating assumptions with a new and more efficient technology.