Fast Accesses to Big Data in Memoryand Storage Systems
1 Fast Accesses to Big Data in Memory and Storage Systems
FastData Accesses in ln-Memory Key-Value Stores2
Fast Data Accesses in In-Memory Key-Value Stores 2
Informationprocessingwithouttables:KeyValueStoresAsimplebuteffectivemethodindataprocessingwhereadatarecord(or a value) is stored and retrieved with its associated keyvariabletypesandlengthsofrecord(value).simpleornoschemaeasysoftwaredevelopmentformanyapplicationsKey-valuestoreshavebeenwidelyused inproductionsystems:GooglefacebookTamazon风LEVELDBTDynamoDBRocksDBLinkedinMProjectVoldemortAdistributeddatabase.redisMEMCACHED
Information processing without tables: Key Value Stores ๏ A simple but effective method in data processing where a data record (or a value) is stored and retrieved with its associated key • variable types and lengths of record (value) • simple or no schema • easy software development for many applications ๏ Key-value stores have been widely used in production systems:
Simple andEasyInterfaces of Key-Value Stores: value = get (key)put (key,value)Variable-lengthkeys&values38John_ageIndexget(key)Client
Simple and Easy Interfaces of Key-Value Stores • value = get (key) • put (key, value) Index John_age 38 . Variable-length keys & values Client get (key)
Key-Value Store:A Data Model Supporting Scale-outCommandOperationGET(key)Read valueSET(key, value)WriteaKVitemDEL(key)DeleteaKVitemHash (Key)>ServerIDItishorizontallyscalableItcanbepotentiallyof(very)highperformance
Key-Value Store: A Data Model Supporting Scale-out Data Servers Command Operation GET(key) Read value SET(key, value) Write a KV item DEL(key) Delete a KV item ▪ It is horizontally scalable, ▪ It can be potentially of (very) high performance. Hash (Key) → Server ID