History[09]

McObject started in 2001 with the launch of the eXtremeDB In-Memory Database System.

Checkpoints

Compression[10]

eXtremeDB offers Zip-like compression for tables (horizontal storage), and Run-length Encoding for time series (vertical storage).

Concurrency Control[11][12]

eXtremeDB offers a multi-version concurrency control (MVCC) optimistic transaction manager and an alternative "pessimistic" MURSIW (MUltiple Reader, Single Writer) transaction manager.

Coordination

Data Model[13][14][15]

Fault Tolerance

Foreign Keys

Indexes[16]

eXtremeDB database offers multiple indexes, including the following:

• B-Trees for common sorting and searches, insertions, and deletions
• R-Trees for geospatial indexing (common in GPS/navigation systems)
• Hash tables for quickly locating a single unique index entry
• Patricia trie index, which speeds searches in networking and telephony applications
• Trigram indexes are ideal for text searches when the exact spelling of the target object is not precisely known. It finds objects which match the maximum number of three-character strings in the entered search terms, i.e., near matches.
• “Custom indexes” or b-trees that allow the application to define the collating sequence of entries; this is useful in implementing soundex algorithm searches, for example
• KD-Trees or k-dimensional trees, for spatial and pattern-matching tasks and in applications where query predicates contain various combinations of object fields (for example, to construct Query-By-Example, or QBE features)
• Approximate Nearest Neighbor (ANN). The native ANN indexes enable high speed similarity searches on high dimensional data directly inside the database engine. The HNSW index is a multi-layered graph-based data structure specifically designed for efficient approximate nearest neighbor search in high-dimensional spaces Vamana index shares conceptual similarities with HNSW, as both utilize graph-based structures for navigation in vector space. However, unlike the multi-layered hierarchy of HNSW, Vamana constructs a single-layer graph (typically based on a directed neighborhood graph, such as NGT or variations of k-NN graphs).

Isolation Levels

Joins

Logging

Parallel Execution

eXtremeDB databases are partitioned horizontally (“sharded”), with each partition/shard managed by an instance of the DBMS server. Shards are typically distributed on a storage array (which may be a SAN) – with each server keeping a CPU core busy – or distributed across different physical servers with their own storage systems. Distributed query processing across multiple servers, CPUs and/or CPU cores accelerates performance. The difference can be dramatic in some cases, with parallel execution of database operations and by harnessing the combined processing power, memory and I/O bandwidth of many nodes rather than just one.

Query Compilation

As well as Prepared Statements

Query Interface

Storage Architecture[17]

eXtremeDB enables the developer to combine both database paradigms – in-memory and on-disk – in a single database instance. Specifying one set of data as transient (managed in memory), while choosing persistent storage for other record types, requires a simple database schema declaration.

Storage Model[18]

eXtremeDB implements row-based layout for all data types other than sequences. Row and columnar layout can be combined in hybrid data designs.

Storage Organization

Stored Procedures[19]

Lua, Python or C++ based stored procedures run in the context of the SQL server and therefore minimizes client-server inter-process communication and attendant network overhead, and fully utilizes the multi-core nature of modern hardware.

System Architecture

User-Defined Functions

Also allows C/C++ and Lua

Views