Concurrency Control

Concurrency Control

1. Why?

   • Lost Updates
   • Uncommitted Data
   • Inconsistent Retrievals

2. When?

   • Multiple Processes
   • Swapping memory
     • From Secondary
       • To Primary

3. Techniques

   • Pessimistic
     • Assumes conflicts will happen
     • Detects conflicts when they happen
       • Resolves them using 'blocking'
     • Sets of locks are used
       • Other users cannot perofm actions that conflict with the lock
       • Owner has to release the lock
     • Uses
       • When there is high contention of data
       • When the cost of protecting data with locks is less than the cost of rolling back transactions if a concurrency conflict occurs
   • Optimisitc
     • Users cannot lock data
     • Raises an error when a user tries to update another user's data
       • The user receiving the error has to roll back the transaction and start over
     • Uses
       • When there is low contention of data
       • When the cost of occasionally rolling back a transaction is lower than the cost of locking data when read

4. What is?

   • Concurrency
     • Computations excuting together and probably interacting
   • Defination
     • Managing simultaneous executions of transactions in a shared database to ensure the serializability
   • Serializability
     • The major correctness criterion for concurrent transactions' executions