Systems Analysis I Data Flow Diagrams

Document Sample
Systems Analysis I Data Flow Diagrams Powered By Docstoc
					           Systems Analysis I
              Data Flow Diagrams


                          ISYS 200
                      Glenn Booker



ISYS 200            Week #4          1
Data Flow Diagrams
    Data Flow Diagrams (DFDs) are the primary
     tool for summarizing the processes a system
     can perform
    They are useful both for capturing the
     existing system‟s capabilities, and then are
     revised to show the new system‟s processes
    A DFD shows processes, entities, and data
     stores, connected by data flow lines

ISYS 200                 Week #4                    2
Data Flow Diagrams
    Processes are the functions some user
     can perform using the system
    Users indicate the kind of user who is
     allowed to perform each process (clerk,
     sales manager, etc.)
    The data stores on a logical DFD are a high
     level description of the types of data storage
     needed to perform each process

ISYS 200                  Week #4                     3
Data Flow Diagrams
    DFDs can show either the logical structure
     of processes, or the physical structure
          Logical structure shows all possible processes,
           regardless of where they occur in the system
          Physical structure indicates what physical parts
           of the system (e.g. type of servers) perform
           each kind of process
    Unless specified, assume we imply a
     logical DFD
ISYS 200                        Week #4                       4
DFD Notation
    Here are Visio
     examples of the                                                   Process

                                                           Label
     entity, process, and                      Entity

     data store shapes,




                                                                       el
                                                                     ab
                                                                   rL
     with data flow lines




                                                                 he
                                                               Ot
     connecting them
                                                          Data Store
          If your shapes differ
           from the text‟s,
                                             Note that in Visio, these shapes
           provide a legend                  are under the Flowchart family,
                                                      not Database
ISYS 200                           Week #4                                       5
DFD Scope
    A DFD does not show any kind of business
     rules for processes – just all possible paths
          No decisions, or conditions (if)
    A logical DFD does not show the subsystems
     involved in a process
          We might show a Customer data store, but
           don‟t specify how or where is it implemented
    Processes may be manual or automated –
     there is no distinction on a logical DFD
ISYS 200                         Week #4                  6
DFD Notation – Entities
    More formally, the „kind of user‟ is called an
     entity in the text – an unfortunate choice
     when we get to the ERD*
           An entity can be a type of user, or some kind of
            external system with which your system interacts
           External systems might include other information
            systems within or outside of your organization
               Like a legacy system used for processing invoices

           * An entity in a DFD could be more general than an entity in an ERD

ISYS 200                                  Week #4                                7
DFD Notation – Processes
    Processes are the business functions
     performed using the system
    We prefer a “verb (adjective) noun”
     naming convention for processes
          Prepare shipping invoice
          Enter new customer
          Remove backorder
          Update reorder parameters
    Number processes for later reference
ISYS 200                      Week #4       8
DFD Notation – Data Stores
    A data store is a kind of data needed to
     perform processes
          Temporary data isn‟t shown on a logical DFD
          Data stores can be manual or automated
    A data store is labeled with the type of data
          Sales
          Customer
    Put a reference number in front of each
     data store name, e.g. D1, D2, D3, etc.
ISYS 200                       Week #4                   9
Creating a Logical DFD
    Data flow diagrams can be done on multiple
     levels of detail, as your understanding of the
     system evolves
          Start by listing the processes and activities
           associated with your system, and look for
           entities, processes, and data stores
          Create a context diagram (the simplest DFD)
          Create the diagram 0 data flow diagram, showing
           entities, general processes and data stores
              Include data flow names

ISYS 200                           Week #4                   10
Creating a Logical DFD
          Pick significant processes, and expand them
           into more detailed DFDs if needed
    The numbering of processes reflects the
     level of detail shown
          The system is process „0‟ in a context diagram
          In a diagram 0 DFD, the processes are
           numbered 1, 2, 3, etc. (parent diagram)
          To expand on process 3, Diagram 3 would show
           processes 3.1, 3.2, 3.3, etc. (child diagram)

ISYS 200                       Week #4                      11
Creating a Logical DFD
    Watch the zeros! Process „0‟ does NOT
     appear on the Diagram 0 DFD!
          Diagram „x‟ means it is a DFD which expands
           on the contents of the process numbered „x‟
          So Diagram 3.2 would contain processes 3.2.1,
           3.2.2, 3.2.3, etc.
    Page 200 has a good Diagram 0 DFD
          This is the lowest level of DFD detail we‟ll use


ISYS 200                         Week #4                      12
Levels of DFD
Context         Shows system as one box, process 0
Diagram

Diagram 0 DFD Shows high level data flows for the
              entire system

Diagram x DFD Shows processes within process
(where x>0)   number „x‟ from the Diagram 0 DFD

Diagram x.y     Shows processes within process
DFD (x, y >0)   number „x.y‟ from the Diagram x DFD
ISYS 200                Week #4                     13
Rules for Data Flow Diagrams
    Entities never connect directly to a data store
          They must use some process to access or modify
           the data
          Hence a user or external system can only connect
           to one or more processes
    All kinds of entities which appear on the
     context diagram must appear somewhere
     on the DFD
          Otherwise they don‟t do anything!
ISYS 200                       Week #4                    14
Rules for Data Flow Diagrams
    Each process will connect to at least one
     user or external system, and one data store
          Each process may send data to one or more
           data stores, and/or get data from one or more
           data stores
          Processes rarely connect to other processes
          Each process needs data flowing in and out of it
          Fix processes which have logically incomplete
           inputs and outputs

ISYS 200                        Week #4                       15
Rules for Data Flow Diagrams
          Leave in processes which calculate
           something, make decisions, manipulate
           data, or organize data
    Data flows into and out of parent and child
     diagrams should be consistent
    Data flows pointing away from an entity
     imply they provide input to the system
    Data flows pointing to an entity imply they
     receive an output from the system
ISYS 200                      Week #4              16
Rules for Data Flow Diagrams
    Every data store needs data flowing both
     in and out somewhere on the diagram
          Otherwise you have a data black hole = inputs
           but no output, or
          A miracle = outputs without input
    Data stores do not connect directly to
     each other
          A process is needed to take data from one
           data store, and write it to another one
ISYS 200                       Week #4                     17
DFD Evolution
    Typically three steps are used with DFDs
          The first DFD captures the processes used
           by the existing system
          The second DFD describes the processes
           to be used by the new system
          A third DFD is a physical description of the
           new system
    The first two DFDs are logical DFDs
          Now we‟ll mention the physical DFD
ISYS 200                        Week #4                   18
Physical DFDs
    A physical DFD shows how processes
     will be implemented
          Entities are still entities
          Data stores are now files, and may include
           temporary files
          Processes are now specific programs or
           manual procedures
    See example on page 202

ISYS 200                       Week #4                  19
Partitioning a DFD
    A Physical DFD can be partitioned to help
     determine the structure of the application
    Partitioning groups processes by similarity
          Group manual processes vs. automated ones
          Group processes by the type of user
          Group processes by their time sequence
          Group processes which perform similar tasks
          Group processes which use similar resources

ISYS 200                      Week #4                    20
Partitioning
          Group processes to ensure consistent data
          Group processes for security protection
    Partitioning can also be done at the user
     interface level (e.g. a web site) for the
     same kinds of reasons




ISYS 200                       Week #4                 21
CRUD Matrix
    A CRUD matrix maps data and processes
     to verify that all data is used correctly
     throughout the system
    CRUD refers to the possible activities
          Create new data
          Read existing data
          Update or change or modify existing data
          Delete existing data


ISYS 200                       Week #4                22
CRUD Matrix
    The CRUD matrix shows process names
     in each row, and data stores in each
     column (p. 203)
          For each process and data store, enter C, R, U,
           and/or D to indicate the possible data activities
          A blank indicates that process does not affect
           that data store
    For further clarity, use the process and
     data store numbers from the DFD
ISYS 200                        Week #4                        23
CRUD Matrix
    At a minimum, make sure each data store is
     Created and Read by at least one process
          Update and Delete may be optional




ISYS 200                      Week #4             24
Event Response Table
    Further documentation of processes is
     encouraged
    One form is the event response table (p. 204)
    This provides a summary of each process
          Event – is the process name
          Source – is the entity who starts the process
          Trigger – is the data or stimulus which is needed
           from the source

ISYS 200                        Week #4                        25
Event Response Table
          Activity – is a brief description of the process
          Response – is typically an interface screen
           (for a human Destination) or type of data (for
           an external system Destination) which
           conveys the results of the process
          Destination – is the entity which received the
           output from the process (whether human or
           an external system)



ISYS 200                         Week #4                      26
Other DFD Documentation
    Another approach for documenting a
     DFD is to outline the contents of the
     DFD before making the drawing
    Identify the entities and processes
    Make a table to show
          Process Name, Description, Entities,
           Inputs, Outputs, and Data Sources



ISYS 200                       Week #4            27
Other DFD Documentation
    Except Description, these all correspond
     directly to stuff which appears on the DFD:
          Process Name – the label inside the process
           shape (later add the numbering)
          Description – a brief description of the process
          Entities – are the entities which participate in
           the process
          Inputs – are data flow(s) leading into the process
          Outputs – are data flow(s) leading out of
           the process
ISYS 200                        Week #4                         28
Other DFD Documentation
          Data Sources – are the data stores connected
           to the process (later add the numbering)
    Notice that the Inputs and Outputs are the
     data flow labels, which could be coming
     from or to entities or data stores
    Use the table to help arrange the processes
          For example, sort the processes by Entity to
           help see which processes should be grouped
           together on the DFD
ISYS 200                       Week #4                    29
Other DFD Documentation
    Duplication of entities and data stores is
     permitted on a DFD for clarity
          Flag or footnote them to indicate a deliberate
           duplicate is being used
    In addition to the description of processes,
     provide a brief description of the data
     stores and entities
          Note any assumptions about the scope or
           responsibilities of anything in the DFD which
           affected your modeling decisions
ISYS 200                        Week #4                     30