notes

Distributed Systems

CAP

1. It’s impossible to achive Consistency, Availability and Network Paritition Tolerance at the same time.

   

Linearizability

1. Linearizability (consistency in CAP) - (simple explanation) when all writes looks like they processed immideately, atomic and when it looks like there is one copy of data.
   • once write completes, all later reads should return the value of that write or the value of a later write
   • once a read returns a particular value, all later reads should return that value or the value of a later write.
2. Properties:
   • no stale reads
   • save state mutation
   • everyone sees the same order of states
3. Serializability (I in ACID) - is a gurantees that the execution of a set of transactions over multiple items is equivalent to some serial execution of the transactions.

Challenges

1. Unreliable network with variable delays
2. Partial failures
3. Unreliable Clocks
4. Processing pauses