NonStop SQL is a database management system designed for 24x7 availability, high scalability, and excellent performance of online transaction processing (OLTP). The availability is achieved by isolation -- the transactions affected by hardware or software failures will be recovered without affecting other transactions. As for scalability, NonStop SQL is able to scale up almost linearly. It's regarded as a base for highly parallel systems. The fast OLTP is due to the row-partition storage model and fine granularity. The lock on data can be at the individual row level. NonStop SQL is widely used in stock exchanges, banks, and telecommunications.
NonStop SQL is the first distributed SQL and the first SQL implementation with fast performance of online transaction processing (OLTP). Previously, all SQL database management systems had a slow OLTP performance, but they were still widely used due to the simplicity of SQL. To compensate, people often used a second, non-SQL database management system for OLTP-heavy applications. The company which developed NonStop SQL -- Tandem -- decided to relieve people from the double database usage. 25 developers of Tandem spent 3 years to write .5 million lines of code as the original code base of NonStop SQL.
NonStop SQL adopted the SQL interface because SQL supports views, assertions, and integrated dictionary. Those were what customers asked for by that time.
Whenever a view is created, NonStop SQL stores a view definition instead of the materialized table. Whenever a user selects data from a view, the data is directly retrieved from the underlying base tables.
Two-Phase Locking (Deadlock Prevention)
The users can set the lock granularity, lock protocol, and lock waits styles in the CONTROL TABLE. The deadlock prevention is based on lock timeouts.
Every single transaction has one single corresponding log record. This log record contains information about how to perform redo, undo, and media recovery.
Nested Loop Join Hash Join Sort-Merge Join Semi Join
NonStop SQL supports hash join, nested loop join, and sort-merge join. The developers of NonStop SQL didn't devote lots of effort on distributed join optimization, because they thought those optimization makes little sense. To the best of our knowledge, NonStop SQL supports semi-join.
N-ary Storage Model (Row/Record)
NonStop SQL uses the N-ary Storage Model, which is appropriate for heavy OLTP workloads.
Read Uncommitted Read Committed Repeatable Read
NonStop SQL supports read uncommitted, read committed, and repeatable reads. Users are able to select isolation levels by themselves.
NonStop SQL supports B+ tree index. By default, the primary key is used as the index of a table. To speed up the non-primary key look up, an index table can be created based on selected columns.
The data of every table is stored in one or more disk files. Each disk file has a label storing the name and catalog information of the associated table.
NonStop SQL supports the basic features of relational database management systems. What distinguish it from many other relational DBMS are the additional features providing excellent OLTP performance, availability, and scalability.
NonStop SQL supports scale-out, and each unit is called a disk volume. Among the disk volumes, the distributed architecture is shared-nothing. In every disk volume, there can be multiple disk servers which are doing data managing, logging, locking, and query execution. Among the disk servers, the distributed architecture is shared-disk.
The query compiler transforms the SQL source commands into executable query plans in host language. The executable query plans and the SQL source commands are stored in a newly created object program file.
Inter-Operator (Vertical) Bushy
For transactions within terabytes of data, the parallelism is mainly in transaction level. Transactions and queries involving more than terabytes of data will be divided into sub tasks and executed in independent disk servers.