Big data types for the Aerospike-architecture series

Large data Types (large datatype)

bins of large data types can store a large collection of objects. The type of the object can be integer/string/map/list. A single record of aerospike can contain many of these bins, which allow a large amount of data to be contained in a separate record, with limits based on storage capacity and no maximum limit for records.

Aerospike the type of storage used by the LDT, we call the child record (sub-records). Child records are similar to Aerospike regular records, except that they are connected to the parent record. The record shares the partition ID and the internal record lock with the parent record, so it moves with the parent record when migrating, and is protected with the same isolation mechanism as the parent record.

Objects in the LDT are not stored continuously with their associated records, but are divided into sub-records (sub-records) (approximately 2kb to 200kb in size, determined by the size of the Write block). The child records are concatenated and indexed together, and the database is internally managed through user-defined functions (UDFs). Using child records means accessing an object in the LDT affects only a few child records, not the entire record.

Large Ordered list (large sequential list)

Large ordered list type (llist) Manage Lists of large-scale sorting. An object in a sequence list can be a primitive type (for example,an integer, a string, a blob) or a complex data type (for example,map, List)

Figure 5a Large ordered list operation

Physically, large sequences are implemented by the B + tree. The nodes and leaves of the tree are child records (sub-records). The root of the B + tree resides in the parent record.

Figure 5b Large ordered list physical layer

There is no depth limit to the B + tree structure, so there is no limit to the number of objects stored in large sequential tables.

The Llist control structure contains all meta-data information about large, serialized tables. For example, include the following information:

• Type of key
• Storage Configuration and Limitations
• The number of roots, nodes, and leaves.
• Storage Status and thresholds
• Array of root and node pointers (b + root node)

Useful Links

• LDT Feature Guide

Original link:

Translator: Beijing It man son

 

Big data types for the Aerospike-architecture series