Please enable JavaScript.

Coggle requires JavaScript to display documents.

Database (Entity Relationship (ER) model (Attributes (characteristic of a…

- - - - lost updates
      - uncommitted data
      - inconsistent retrieval or incorrect summary
    - - Shared lock (s lock)
      - exclusive lock (x lock)
      - Binary lock
        
        1 locked 0 unlocked
      - Two phase locking protocol (2PL)
        
        works by delaying unlocking till the end of the first transaction
        
        issue
        
        does not ensure deadlock-freedom
        
        deadlock occurs when two transactions wait for each other to unlock a shared object
        
        concurrency controll (through timestamps ans transaction rollback)
        
        wait-die scheme
        
        2 more items...
        
        wound-wait scheme
        
        2 more items...
        
        ensures serialisability ( so deals with concurrency anomalies)
  - - - the execution of a transaction in isolation (ie. with no other transaction executing concurrently) leaves the database in a consistent state
    - - a transaction should appear as though it is executing in isolation; that is, each transaction is unaware of the other transactions executing concurrently in the system
    - - a transaction is an atomic unit of processing; either all operations are performed or none are
    - - when a transaction completes successfully, the changes it made on the database persist, even if there are system failures
  - - - if at any point during execution the transaction fails and has to be aborted:
      - delete new copy and point back to old copy
    - - if the transaction has been committed(completed its execution) then all updates have been performed on the new copy of the db
      - db pointer set to new copy and delete old copy(original copy aka shadow copy)
- - - - services rendered
      - client payments recieved
      - product produced
      - payments made to suppliers
      - order taken
    - - Access to Historic Data
      - Data in multiple stand -alone OLTPs
      - OLTP System design - intention
      - hence to avoid this
        
        take info needed for BI to a different data store outside OLTP systems
        
        data mart
        
        Dimension
        
        Attribute
        
        Measure
        
        Hierachy
        
        OLAP Online analytical processing
        
        enables user to (quickly and easily) retrieve information from data (usually in a data mart) for analysis
        
        designed to let the user interact with data ( slicing and dicing the data to view it )
        
        drilling down
        
        1 more item...
        
        rolling up
        
        1 more item...
        
        built around measures and dimensions to present data as a multi dimensional structure (cube)
        
        architectures
        
        HOLAP low latency med scalability mid-high speed
        
        MOLAP high latency low scalability high speed
        
        ROLAP low latency high scalability low speed
        
        Alternative
        
        UDM (unified dimensional model)
        
        3 more items...