On Github ukmadlz / nosql-is-a-lie-devtalks-2016
IBM Cloud Data Services
try { SQL } catch (Exception $e) { Must be NoSQL }
It's a nasty backronym
- Reintroduced by Johan Oskarsson of Last.fm in 2009 to describe all upcoming DB techSPARQL
CQL (Cassandra Query Language)
Couchbase SQL
Yes, they'll accept anything…but
Because how else do you know what you’re getting out?
So many distinct types of databases
In = Out
Know's where to look
Ignoring the Partition Tolerance by being in the same place
Needs to know what machine X data is on
The reason why most NoSQL Databases are either AP or CP
Key-Value
Document
Column
Graph
Object, Tabular, Tuple, Triple/Quad store (RDF), Multimodel, Etc
Different indexing & lookups on the same stored data.
Redis, Memcached, Riak KV, Hazelcast, Ehcache, Aerospike, Oracle Coherence, Berkeley DB, Amazon SimpleDB, Oracle NoSQL, Infinispan, LevelDB, GridGain, ZODB, GT.M, NCache, RocksDB, WiredTiger, WebSphere eXtreme Scale, Tokyo Cabinet, Project Voldemort, XAP, Hibari, MapDB, Tokyo Tyrant, STSdb, Scalaris, GlobalsDB, HyperDex, Kyoto Cabinet, Tarantool, LightCloud, ScaleOut StateServer, Upscaledb, Quasardb, Bangdb, BergDB, Cachelot.io, CodernityDB, CortexDB, Elliptics, Helium, HyperLevelDB, Kyoto Tycoon, LedisDB, Nanolat, Resin, TomP2P
Cloudant, CouchDB, MongoDB, Couchbase, RethinkDB, RavenDB, GemFire, PouchDB, Microsoft Azure DocumentDB, Datameer, CloudKit, Mnesia, Google Cloud Datastore, TokuMX, Clusterpoint, Terrastore, RaptorDB, EJDB, SisoDb, WhiteDB, Sequoiadb, JasDB, LokiJS, DensoDB, Djondb, FaunaDB, FleetDB, SenseiDB
Cassandra, HBase, Accumulo, Hypertable, Google Cloud Bigtable, ScyllaDB
Neo4j, Titan, Giraph, InfiniteGraph, Sparksee, HyperGraphDB, FlockDB, VelocityGraph, InfoGrid, GraphBase
NoSQL is a lie
Why NoSQL grew up the way it did
And it's just different databases for different jobs
Some reasons to use different databases