Advantages and limitations of key-value databases

NoSQL Concepts

Miriam Antona

Software engineer

Advantages - very simple

• Key-value tuple
• No defined schema/types
• Basic operations:
  • Put
    • inserts a new key-value tuple
    • updates a value if the key already exists
  • Get
    • returns the value by a given key
  • Delete
    • removes a key and its value
• Fast operations

Advantages - flexible

• Allow changes in data types
  • userID:123 = 123456
  • userID:123 = "Miriam"
• Add additional attributes
  • user:457:preferences = {"language" : "en:US"}
  • user:457:preferences = {"language" : "en:US", "color" : "green","timezone" :"GTM-4"}

Advantages - information stored in memory

• Fast reads/writes
• Can lose data
• Combination of disk and memory persistence

Advantages - scalability

• Can scale horizontally
• Sharding
  • distributes different parts of the data across multiple servers

Advantages - scalability

• Can scale horizontally
• Sharding
  • distributes different parts of the data across multiple servers

Advantages - scalability

• Can scale horizontally
• Sharding
  • distributes different parts of the data across multiple servers

Advantages - scalability

• Can scale horizontally
• Sharding
  • distributes different parts of the data across multiple servers

Limitations

• Just search by key
  • Problem if we don't know the key
  • Some key-value databases added functionalities
    • search by value
    • add secondary indexes
    • search by several keys simultaneously
  • Not complex queries

Let's practice!

NoSQL Concepts