If you are trying to decide what kind of database to pick for your project, the following principles and theorems would be beneficial to help you make that decision.

ACID

• Is a set of properties of database transactions intended to guarantee validity even in the event of errors

• Atomicity
  • Each transaction be “all or nothing”
  • To the outside world: a committed transaction appears to be indivisible, and an aborted transaction does not happen
• Consistency
  • Any transaction must change affected data only in allowed ways
  • All data written to db must be valid according to all defined rules
• Isolation
  • Defines how/when changes made by one operation become visible to other
• Durability
  • Once a transaction has been committed, it will remain so, even in the event of power loss, crashes or errors

CAP Theorem (also named Brewer’s Theorem)

• It’s impossible for a distributed data store to simultaneously provide more than two out of the following three guarantees:
  • Consistency - every read receives the most recent write or an error
  • Availability - every request receives a (non-error) response - without gurantee that it contains the most recent write
  • Partition tolerance - System continues to operate despite an arbitrary number of messages being dropped or delayed by the network between nodes
• OR - stated simply:
  • In a network partition, one has to choose between consistency and availability
  • Because:
    • No distributed system is safe from network failures -> network partitioning generally has to be tolerated
    • Hence, left with: consistency or availability

DB systems designed with traditional ACID such as RDBMS choose consistency over availability

PACELC theorem

• Is an extension to CAP theorem
• Both theorems describe how distributed databases have limitations and tradeoffs regarding consistency, availability, and partition tolerance.
• PACELC goes further: A trade-off also exists, between latency and consistency (even in the absence of partitions)