Database design considerations

Similar Questions

• database design patterns

  I would like to learn more on "database design patterns ', on the internet, I can't find good web page for this.
  It should describe what data models are performed all list and describe when to use them.

  CharlesRoos wrote:
  Design patterns traditional relational database transactional table is what I'm looking for.

  There is not really such a beast - or at least no I ran into when dealing with RDBMS designs. The approach that is usually used is a "high level data model" (often called an entity-relationship diagram). This DR. would then become the plan of action for the implementation of the physical database model.

  Also such patterns exist in the database. Simpler is so-called ' header and lines plan tables "which should explain to create 2 tables and link/join with next path and it can become such performance or other problems with this model and so on.

  This isn't really a model - is standardization and what is expected of relations in 3NF.

  Also, it is wrong to assume that a "+ header - and - line-tables +" exist purely for performance problems. In fact, it can actually be slower done as 2 separate table (using 3NF) more simple using only one table (2NF).

  I remember a design methodology called Tetrach (not sure of spelling) I have done a few courses in, in the 1980s. It deals with the prediction of the performance of the relational data models. Simply put - if you have an invoice that usually consists of a single header line and 20 lines of line, processing invoices will require an average of 21 St / s (logical or physical) by Bill. 2NF, it will take a single e/s - however, the data model and then imposes a limit of 20 lines per invoice (never forget the story on such a structure coding problems).

  However, if the system is I/O forcing then? You decide on a single i/o per invoice (by limiting the flexibility of the data model) or 21 St / s per invoice (and face the real possibility that the system may be impossible to successfully treat all invoices received in time for shipment orders)?

  Well, this isn't really that relevant more with modern equipment of today and storage systems - but it still illustrates to the point that performance alone is not the primary consideration for database designs.

  When to use patter 'create 2 separate identical tables to separate the same kind of business objects, for exmple GoodTeachersTable and BadTeachersTable' and explain why and when creating this model. And so on. There are 30 such techniques of model I guess than that. University teaches only that we have "one-to-many" - thing in the world, that I'm not interested.

  I think you're confusing models of object oriented with relational data modeling design. There is no "+ patterns +" in the data model relationship - with the exception of different normal is a get - with the fundamental objective of correctly normalizing the data model to 3NF as part of a proven recipe of a robust, functional and scalable data model design.

• OLTP database design...

  Dear team,
  
  I would like to know some books / links that explains the concepts of OLTP database design...
  
  If one of the experts has done this before, please share experiences and ideas...
  
  Thank you
  NIC...

  Thanks for the correction,

• Why sql sever database design is used for essbas?

  Hello
  I'm new on Hyperion technology. I would like to know why we use any RDBM as sql sever design of database for essbase, which in itself is the database. What is the main difference between these two databases?
  My understanding is that law Essbase as same as SSAS for cube and the strengthening of the dimension. And they use the database design to store tables and other information.
  
  Thank you

  A relational database can be used for a number of reasons with Essbase (and in some cases). It comes
  1. a repository for system tables for Shared services, EAS, Studio, EPMA, workspace, etc. of financial management
  2. a data source to load Essbase
  3. a workspace for transformations

• Database design issues

  Y at - it a forum that deals with issues of General database design, such as the definition of primary keys, unique constraints, Check constraints, indexes, etc.?
  I searched but could not find it. Maybe I missed it. Thank you.

  General, database or SQL/PLSQL:

  http://forums.Oracle.com/forums/category.jspa?categoryID=18

• Database design

  What are the different approaches to solve this problem?

  Design of database tables to mimic the hierarchical file system.  (e.g. drives, directories, files, include common date for properties such as name, readonly, size, created).

  (a) name of the database 'Interview_FirstnameLastName'.

  b) Design should support files in the root of the drive, or any number of levels down.  (for example c:\test.txt, d:\test\text.txt, c:\test\test\text.txt, etc.)

  c) create a script to insert test data in tables.  Must have at least 25 records.  You can also create processes to take a base path and fill the database by browsing a directory tree.

  (d) provide scripts to create the tables, relationships, indexes or a backup of the database. 

  e. write a stored procedure (usp_FileSearch) that takes the text 1 (@criteria) parameter.  It should return a containing all defined data files whose name contains @criteria.   Returns the ID of the file, the full path, file name, readonly, size, date of creation)

  Frank did not mention explicitly (yet)

  A tree (data structure) perfectly fits your needs:

  1. root = drive
  2. nodes = files
  3. sheets = files

  Concerning

  Etbin

• Oracle database design

  Hello
  
  I am designing a database and want to design it inspires 3NF. I have a table and columns that are not a ship a lot of relationship. I can design so the columns that are not a child either in a few separate tables, with a one to one relationship? Example: My main table is a table of program, and I have a few columns DEsc, Date of analysis analysis and analysis BY. I want to keep these columns in a separate table. Please let me know how to draw this?
  
  Thanks for your time and your help.

  Hello

  ssk1974 wrote:
  ... I currently have the program tables and detailed program with many ships of relationship. ... Can I do the columns of analysis as a one to one relationship and place them in a separate table? How I who would design.

  Of course; You can do it. Let's say the primary key of the program is program_id, and that there is also a program_id column in the program_detail which refers to program.program_id.

  If program_detail contains only the columns 'analysis', then you can simply modify program_detail.program_id to make it unique.

  If program_detail contains other columns, and the relationship with the program must be one-to-many for them, then create another table (let's call it program_analysis), where the primary key is program_id, who is also a foreign key referencing program.program_id. Copy all the columns in the analysis of program_detail to program_analysis, and then move the columns in program_detail.

  As mentioned earlier, make sure you have a good reason to do this first. Normally, everything that has a one-to-one relationship with program_id belonged to the table in the program. There may be some benefits for some of the columns in a table separate program_analysis, but there are costs, too. Make sure that the costs are lower than the benefits.

• How to apply the physical database design

  I imported the tables in my physical database and made some design changes. Now how can I apply the changes to my physical database?
  
  On the Tools menu, patterns for comparison/merging of views can make the comparison between registered design but no physical database?
  
  Published by: 871744 on April 10, 2012 20:31

  In the right part of the main toolbar, there are two blue arrows for synchronization to/from the data dictionary.

• AS3 + mysql database design: several students over several results

  It is more a problem of design of database, but almost everyone here using database integration with their applications so it seems important. I'm integrating AS3 + mysql + php and need to clarify the design of database of several results of students in various activities.

  Development of a learning course online with a 'dashboard' control panel with all the kids teacher translates every activity.

  I read the rules of design of database by Codd etc...

  I understand the need for the parent tables and children so information are relational and not duplicated etc...

  BUT I'm not sure I know how to design the results section. Several students various lessons with 10 activities which each give a mark out of ten.

  The table would be updated every few seconds that 10,000 users would be online at the same time

  1. I created a table of results with the results of pupils in the same table, IE: Tom would be the first line

  brand name lesson activity

  Tom               1                1               6

  John              1                1               4

  Tom               1                2               8

  As you can see it Tom is very fast, because he has already completed the activity 2 and so it is added to the third rank. It seems I'm repeating the same info: lesson and Tom.

  I know that this should be a student_id instead of the name of Tom - it was just to clarify the example.

  Thank you

  Could someone send me to a link with an example of a table for the results of the students

  nothing is repeated in this table (results).  you have a number of lessons each with 10 activities.  It's no different than having 10 times more lessons.

• relational database design

  I apologize in advance for the banality of this question, I have been the design of Foundation for my business for years and I am comfortable with html and recently designed a site complex havig susssed .css and coldfusion code written however dw.
  I CANNOT MAKE THE LEAP TO GET MY HEAD around RELATIONAL DATABASES and want to integrate some forms of dynamic application (which I can code in CF).
  Can someone point me to a resource that will walk me but how to build a simple relational data base (then a complex) that send me to sleep or make sure my eyes glaze.
  Its a brain block which now cause me stress, but I just don't understand...
  Save me
  
  Ta

  No worries. I can sympathize with trying to grasp new concepts. problem is that double signage makes just more difficult for people to follow the wire (s).

• Simple database design - of the Suggestions?

  I built a sample non-real life data base for an auto Agency. The Agency has, model cars, some sellers and buyers. I have to keep track of each of them.
  
  Here's the process I took during the design:
  Standards 1, 2 and 3NF, and relationship to an and the other to many...
  
  Column of italics is primary key...
  
  The database is not realistic, I did it for the practice of the purpose. If you please, look at the diagram and tell me do you think its OK so far? or the design is wrong?
  
  This is my schema:
  
  Vendor table stores the information from the Publisher:
  
  * Seller Table {+ SellerID + (number), first name (text), family name name (text), address (text), city (text), CountryID (foreign key), hired to date (date)} *.
  
  Table Customer stores customer information, the client can deal with a seller at a time, while a seller can deal with more than one customer at a time:
  
  * Customer Table {+ CustomerID (number), first name (text), name of family (text), address (text), city (text), (foreign key) CountryID, SellerID (foreign key)} *.
  
  Car table stores information about the cars available:
  
  * Car Table {+ CarID (number), manufacturer (text), model (text), car year (text)} *.
  
  Bought car table hold information about customers who bought a car and a car of type were purchased and the date of purchase (date of purchase is based on CarID and CustomerID field):
  
  * Bought car Table {CarID + (foreign key) + CustomerID (foreign key), purchase date (date)} *.
  
  OR
  
  * Bought car Table {+ PurchasedCarID + (number), CarID (foreign key), CustomerID (foreign key), date (date of purchase)} *.
  
  -----
  
  * Country table {CountryID + (number), name of the country (text)} *.
  
  A customer and a seller can store several phone number:
  
  * Phone of Table {PhoneID + (number), CustomerID (foreign key), phone number (text)} *.
  
  If I want a salesman to store more than one number, should I do a new table that stores the sellers phone numbers? Or there is a better way to do?
  
  Note: I created country table and table of phone number because they are considered as multivalued fields. To achieve 1NF, I must leave each field in the table (e.g. seller and customer) be atomic.

  It would be better if you post real DDL for the proposed tables.

• table of database design

  Hello
  I have for the design of database in the oracle database.
  I'm designing tables
  Here, I have a problem I have to store the following data
  can someone help me in this metter how to design or tables for these data. acually these data are very important for my database to view the emploee performance
  
  EMPNO Communication Application_knowledge Programming_skills Team_work Analytical_ability Architecture
  E001 quite satisfactory satisfactory very satisfactory needs to improve
  E002 strong needs satisfactory improvement strong very satisfactory
  Very strong satisfactory E003 needs satisfactory satisfactory improvement
  E004 should improve very satisfying loud very satisfactory
  Strong E005 should improve very satisfactory very satisfactory
  E006 strong strong strong strong strong strong
  E007 satisfactory satisfactory very satisfactory satisfactory satisfactory
  
  
  Thank you

  Hello

  ID of the jurisdiction would be the primary key (it must be a look up table) for more details Emp.

  -Pavan Kumar N

• Database design question

  Hello, I have a one-to-many relationship problem when designing a store management system. Each store can have several names, but there at least a name. Any suggestion? Thank you

  Try this:

  CREATE TABLE MyStores

  (

  --E t c...

  , Store_Name_Id VARCHAR2 (7) NOT NULL

  CONSTRAINT Fk_Store_name

  REFERENCES Store_Names_Tab (Name_Id)

  -- ...

  )

• Design considerations

  Hello

  I'm planing on 5 hosts in a cluster deployment, and I think to the following:

  (1) is it recommended that all hosts are runing the same configuration (memory, NIC, CPU, etc.)?

  (2) is recommended to have only one supplier? (Dell, or HP) or better to mix?

  (3) is there anywhere I can check the best compatible with ESXi 5 material? An official site?

  Can you please provide me with reference links to the answers in addition to your own experience with justification if possible?

  Kind regards

  Most recommend against it.  You'll end up with different configs for different types of hosts: several configurations, the firmware version, sellers to manage.

  I don't see a good reason to mix vendors.

• Database design - M:M

  Hello friends,
  I'm confused, how do many many relationships in my application:
  
  I have agents and projects of two tables.
  Each agent can manage several projects.
  Each project can be handled by several agents.
  
  Example of table INIT_PROJECT:

  ID , PROJECT
  1  , X Tower
  2  , Y Tower
  3  , Z Tower

  Example of table INIT_AGEN:

  ID , Agent
  1  , A
  2  , B
  3  , C

  What is the correct way to join or to represent the relationship between the two tables to get a table that servers as a lookup table also:
  
  is - it to be like this:
  
  Example 1 in new_table
  

  Project_ID , Project_name, handled_by1, handled_by2, handled_by3, handled_by4
  1             , X tower        ,  A              , B

  or is it:

  ID            , Project_name, handled_by
  1             , X tower        ,  A            
  2             , X tower        ,  B              
  3             , X tower        ,  C   
  4             , Y tower        ,  A              
  5             , Y tower        ,  B              
  ........
  .....             
             
    

  Then, on a FORM I have lists of values.
  Projects and Agents.
  When the user selects a project, he or she can only see the agents engaged in this project in the list of Agents of values.
  I use a shuttle box for inserting the values "managed by".
  
  Or do like this:
  Example 2 of new_table

  ID , PROJECT,...,HANDLED_BY
  1  , X Tower,... ,A:B:C:D
  2  , Y Tower,... ,A:F:G:I
  3  , Z Tower,... ,F:A:I:W

  I have two lists of values:
  'Project name' and 'Agent '.
  When the user selects a project, then he must see only the values of the handled_by that are associated with this project only.
  
  I used this query to prepare the values in the list of officers of the value
  

  SELECT  
    distinct REGEXP_SUBSTR(handled_by, '[^:]+', 1, LEVEL)  d , REGEXP_SUBSTR(handled_by, '[^:]+', 1, LEVEL) r
     
    FROM  (
           SELECT   project, ROWNUM AS id
           ,      handled_by
           FROM   projects 
          ) where project = :P8_project
    CONNECT BY INSTR(handled_by, ':', 1, LEVEL-1) > 0
           AND id = PRIOR id
           AND PRIOR DBMS_RANDOM.VALUE IS NOT NULL; 

  Can you please advice what is the right way?
  
  Best regards
  Fateh

  organize projects of the agent in another table, which means that you should have 3 table, the agents, the projects and the agent_projects.

  ID            , Project_name, handled_by
  1             , X tower        ,  A
  2             , X tower        ,  B
  3             , X tower        ,  C
  4             , Y tower        ,  A
  5             , Y tower        ,  B   
  

  for an OLTP system, of course.