Better Primary Key

What will make a better primary key? What is the best primary key Data type?

If we're talking about synthetic primary keys, I'll agree with you, Andrew. Much easier to create a synthetic key using a NUMBER column.
If we're talking about natural keys, though, I'm not so sure. If a VARCHAR2 column is truely the unique, immutable identifier for a row-- understanding the semantics of VARCHAR2 uniqueness, perhaps you really want case sensitivity-- I see no reason not to use that column as the primary key.
Justin
Distributed Database Consulting, Inc.
http://www.ddbcinc.com/askDDBC

Similar Messages

EF beta - views with no primary key

When importing views into the entity framework designer, it often gives this error (in the edmx file only; no error is shown in the UI):
warning 6013: The table/view '[view name]' does not have a primary key defined and no valid primary key could be inferred. This table/view has been excluded. To use the entity, you will need to review your schema, add the correct keys, and uncomment it.
I don't want to cloud the issue by going into the details of the views in question; there will always be views where no PK can be inferred automatically, for many reasons. And I am quite happy to define the key fields myself after import. The problem is that it can't be imported at all. So my question to the ODP.NET developers is: can you just import it, and leave me to define the PK?
The workaround I am using is:
1) define a temp table of the same structure
2) temporarily recreate the view as "select * from [temp table]"
3) update the model from database
3) recreate the proper view
But it's a pain. I have to remember to do that every time I update the model from the database. If I forget, it wipes the mapping, and I have to redefine it manually, field by field.

Unfortunately there is still a problem, in a minority of cases:
Error 75: Key Part: '[field name]' for type [view name] is not valid. All parts of the key must be non nullable.
This happens when the view field is not marked "not null". Generally it flows through from the underlying table field. But one of my views involves a "union all". The underlying table fields are marked "not null" but it doesn't make it through the "union all". The view has a primary key, but still the error.
I would rather it just trust me that primary key fields will never be null.

Atomic Primary Key or Composite ? Which approach is better ?

Hi all,
I have a table that has its composite primary key (infact foriegn keys of other tables make pk in this table).
But due to large project complexity, i just want to have an atomic primary key, so that i could use a sequence for it, instead of using N foreign keys from other tables. And furthermore, a single column primary key can be easily accessed as foreign key as compared to N columns primary key.
Is it a good approach in general ?
My personal view has certain reservations regarding the atomic primary key in such case. It could lead to a weak database design.
Any suggestion ?
Thanx
Zaaf.

Most large data warehouses implement a 'synthetic' key approach for much the same reasons that you describe. You can still carry the FKs and use them to assist in joining operations, while keeping the size (and complexity) of the primary key down to a manageable size.
There are downsides to this approach - it depends on how the data is most frequently accessed and many other factors. But in general it can work quite well.

Primary key index not working in a select statment

Hi ,
I'm doing a :
select *
from my_table
where B = v1 and
A = v2 ;
where A and B are the primary key of the table, in that table the primary key index is there and the order of the primary key definition is A first and then B.
While testing this statment in my database the Explain Plan shows that it is using the index ok but when
runninng in client database it is not using it becasue of the order (i think), is this something configurable that I need to ask to my DBA ?
The solution I found was to do the select with a hint wich fix the problem , but I'm curious about why is working in my dadabase and not in the client database
thanks in advance .
oracle version 11g

This is the forum for SQL Developer (Not for general SQL/PLSQL questions). Your question would be better asked in the SQL and PL/SQL forum.
Short answer: The execution plan used will depend on optimizer settings and table/index statistics. For example if the table has very few rows it may not bother using the index.

Application Translation Not Working - Primary Key Error

I had created an application translation to Spanish but it wasn't displaying my Spanish translation. I was going to try and redo it so I tried creating a new mapping and then seeding the translatable text. I got a primary key error "ORA-20001: Seed insert error: WWV_FLOW_TOPLEVEL_TABS.TAB_TEXT ORA-00001: unique constraint (FLOWS_030100.WWV_FLOW_TRANSLATABLE_TEXT_PK) violated".
I went in and deleted what I had done through APEX for this app and tried to create a new application but still get the primary key error when I try to seed it. I gave it a new translated application ID but that doesn't seem to help. Anyone have this problem or no of the reason I'm having this issue?
Thanks.

Hi David,
Thanks for reporting this. This was an interesting problem to solve.
As it turns out, it was a logic error in Application Express (i.e., bug). When deleting a translation mapping, the associated strings in the translation repository would be deleted for that application, but if and only if you had actually published the application.
I think this is why you could never get Spanish working properly - you had never actually published the application the first time. So I'll bet what you did is deleted the original mapping, then you recreated the mapping for the same language but with a different translated application ID. Since you had never published the application from the original translation mapping, and there were orphaned rows in the translation repository, you encountered a "collision" when you tried to seed the second time with the different translated application ID.
Your action of deleting rows from WWV_FLOW_TRANSLATABLE_TEXT$ cleaned up these orphaned rows. As you stated, this isn't recommended to perform DML on the underlying APEX tables. A couple alternatives could have been:
1) Before deleting the translated application mapping, actually publish the application and then delete the mapping.
2) If you had deleted the mapping already, you could recreate the mapping for the same language and with the original translated application ID. Then, publish the application and then go back and delete the mapping.
I realize all this sounds crazy. But it was only an issue because you had not actually published the application. Not your fault, though, as this is a bug in APEX.
This bug will be fixed in Application Express 4.0. This way, you won't have to worry about if you published or didn't publish. The orphaned rows will be cleaned up when you delete a mapping.
Thanks again for reporting this.
Joel

How to create a primary key

Hello - New to Oracle and was wondering what is a good way to create an auto-incrementing primary key for a table. I have done this so far:
Created a new sequence cst_swq begin with 1 increment by 1.
CREATE OR REPLACE TRIGGER CST_TRIG
BEFORE INSERT
ON COOKIE_SALES
REFERENCING NEW AS NEW
FOR EACH ROW
BEGIN
SELECT CST_SEQ.nextval INTO NEW.SALE_ID FROM dual;
This does not really work for me. Help with the code or a better idea to create primary key is much appreciated, thank you.

adi26 wrote:
This does not really work for me. What does that mean?
In which way does this approch not meet your requirements?
BTW: after 200+ posts you should have learned how to use tags...
bye
TPD

Parameters in Url..how do I use login information to insert the primary key

Uggg, it is 3 am, I am so tired. Please help.
I have made a registration page for my site. Hurray it works.
Login Page. It works
Protected some pages. Works
Now I am trying to make a page that the users can go to and update their user information. User name, password, email ect.
I can get it to prepopulate if I put ?id=3, or someother primary key in the url by hand.....and it does update the data.... but I can't, for the life of me, figure out how to make it use the users login information to get and insert the primary key in the url from the link to the update page.
I have used the hyperlink box and chose the updateaccount.php file, then went to parameters and selected Name: id Value: I chose id from the dropdown from the data base. This puts in an echo that looks like it should work. NO SUCH LUCK.
Pleeeese!!! What am I doing wrong.
I am a total novice. Done a few static sites. So please speak slowly.
Thanks sooooo much

I really can´t recommend exposing critical stuff like e.g. the user´s login_id in the URL, because once it´s visible, it can be changed by anyone. Better insert the Session Variable kt_login_id, which will only be availabe on a custom PHP page when inserting...
<?php session_start(); ?>
...on line 1 of this document.
Cheers,
Günter

Entity bean with another class as primary key generator class

Hi All,
I have a CMP entity bean in which I generate the primary key using my own class know as unique id generator.
I use it in ejbCreate method in the following manner
public Long ejbCreate(HospitalData hospitalData) throws CreateException{
          Long myid = new Long(UniqueIdGenerator.getId());
          System.out.println("My id generated ====== "+myid);
          this.hospitalid = myid;
          System.out.println("My id generated ====== "+this.hospitalid);
          System.out.println("Came in ejbCreate method of HospitalEJB");
          this.hospitalname = hospitalData.getHospitalname();
          return null;
Can you tell me how I map this primary key in my ejb-jar.xml and jbosscmp-jdbc.xml.
Anyhelp would be appreciated.
Thanks
Sameer

"Bhamu" <[email protected]> wrote in message
news:9keuo4$[email protected]..
I am trying to develop an entity bean attached to a key which have a
composite key using Primary Key Class. When I use the findByPrimaryKey
method of the bean it throws an exception as follows,I notice that you are using the reference CMP plugin.
at com.netscape.server.ejb.SQLPersistenceManager.find(Unknown Source)I'm also willing to bet that you may have created your own primary key class
with a single field, rather than using the fields type as a primitive
primary key class. If you do this, then SQLPersistenceManager will break.
It is badly written, and has some stupid assumptions. The one I remember
the most is that if you only have one primary key field, then
SQLPersistenceManager assumes you have used a primitive type such as
java.lang.String or java.lang.Integer, to represent it, rather than creating
your own <Enity>Pk.java file.
SQLPersistenceManager works for toy examples, but in general I would say it
is broken and unusable. Either use BMP, or splash out the money for Coco
Base from Thought Inc. Currently the only CMP plugin for iPlanet App server,
other than the reference implementation, that I know of.

Entity Bean with Primary Key Class

Hi
I am trying to develop an entity bean attached to a key which have a
composite key using Primary Key Class. When I use the findByPrimaryKey
method of the bean it throws an exception as follows,
java.sql.SQLException: Parameter typecheck failed: invalid parameter type:
ORA-00000: normal, successful completion
at com.netscape.server.jdbc.PreparedStatement.verifyError(Unknown Source)
at com.netscape.server.jdbc.PreparedStatement.afterVerify(Unknown Source)
at com.netscape.server.jdbc.PreparedStatement.executeQuery(Unknown Source)
at com.netscape.server.ejb.SQLPersistenceManager.find(Unknown Source)
at
com.netscape.server.ejb.EntityDelegateManagerImpl.doPersistentFind(Unknown
Source)
at com.netscape.server.ejb.EntityDelegateManagerImpl.find(Unknown Source)
at com.kivasoft.eb.EBHomeBase.findSingleByParms(Unknown Source)
at
TestBean2.ejb_home_TestBean2_EntityBean2Bean.findByPrimaryKey(ejb_home_TestB
ean2_EntityBean2Bean.java:120)
at
TestBean2.ejb_kcp_skel_EntityBean2Home.findByPrimaryKey__TestBean2_EntityBea
n2__TestBean2_TestBKey(ejb_kcp_skel_EntityBean2Home.java:258)
at com.kivasoft.ebfp.FPRequest.invokenative(Native Method)
at com.kivasoft.ebfp.FPRequest.invoke(Unknown Source)
at
TestBean2.ejb_kcp_stub_EntityBean2Home.findByPrimaryKey(ejb_kcp_stub_EntityB
ean2Home.java:341)
at
TestBean2.ejb_stub_EntityBean2Home.findByPrimaryKey(ejb_stub_EntityBean2Home
..java:88)
at TestBean2.Simple1.defaultAction(Simple1.java:115)
at TestBean2.Simple1.doGet(Simple1.java:26)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:740)
at javax.servlet.http.HttpServlet.service(HttpServlet.java:865)
at com.netscape.server.servlet.servletrunner.ServletInfo.service(Unknown
Source)
at com.netscape.server.servlet.servletrunner.ServletRunner.execute(Unknown
Source)
at com.kivasoft.applogic.AppLogic.execute(Unknown Source)
at com.kivasoft.applogic.AppLogic.execute(Unknown Source)
at com.kivasoft.thread.ThreadBasic.run(Native Method)
at java.lang.Thread.run(Thread.java:479)
javax.ejb.FinderException
at com.netscape.server.ejb.SQLPersistenceManager.find(Unknown Source)
at
com.netscape.server.ejb.EntityDelegateManagerImpl.doPersistentFind(Unknown
Source)
at com.netscape.server.ejb.EntityDelegateManagerImpl.find(Unknown Source)
at com.kivasoft.eb.EBHomeBase.findSingleByParms(Unknown Source)
at
TestBean2.ejb_home_TestBean2_EntityBean2Bean.findByPrimaryKey(ejb_home_TestB
ean2_EntityBean2Bean.java:120)
at
TestBean2.ejb_kcp_skel_EntityBean2Home.findByPrimaryKey__TestBean2_EntityBea
n2__TestBean2_TestBKey(ejb_kcp_skel_EntityBean2Home.java:258)
at com.kivasoft.ebfp.FPRequest.invokenative(Native Method)
at com.kivasoft.ebfp.FPRequest.invoke(Unknown Source)
at
TestBean2.ejb_kcp_stub_EntityBean2Home.findByPrimaryKey(ejb_kcp_stub_EntityB
ean2Home.java:341)
at
TestBean2.ejb_stub_EntityBean2Home.findByPrimaryKey(ejb_stub_EntityBean2Home
..java:88)
at TestBean2.Simple1.defaultAction(Simple1.java:115)
at TestBean2.Simple1.doGet(Simple1.java:26)
[03/Aug/2001 15:26:41:7] error: Exception: SERVLET-execution_failed: Error
in executing servlet Simat
javax.servlet.http.HttpServlet.service(HttpServlet.java:740)
at javax.servlet.http.HttpServletple1: java.lang.ClassCastException:
com.netscape.server.deployment.EntityBeanDescriptor
Exception .service(HttpServlet.java:865)
at com.netscape.server.servlet.servletrunner.ServletInfo.service(Unknown
Source)
at com.netscape.server.servlet.servletrunner.ServletRunner.execute(Unknown
Source)Stack Trace:
java.lang.ClassCastException:
com.netscape.server.deployment.EntityBeanDescriptor
a
at com.kivasoft.applogic.AppLogic.execute(Unknown Source)
at com.kivasoft.applogic.AppLogic.ext
com.netscape.server.servlet.servletrunner.AppInfo.popWebAppDescriptor(Unknow
n Source)
at com.netscape.server.servlet.servletrunner.ServletInfo.service(Unknown
Source)
at com.netscape.server.seecute(Unknown Source)
at com.kivasoft.thread.ThreadBasic.run(Native Method)
at java.lang.Threadrvlet.servletrunner.ServletRunner.execute(Unknown
Source)
at com.kivasoft.applogic.AppLogic.execu.run(Thread.java:479)
te(Unknown Source)
at com.kivasoft.applogic.AppLogic.execute(Unknown Source)
at com.kivasoft.thread.ThreadBasic.run(Native Method)
at java.lang.Thread.run(Thread.java:479)
Any kind of help is appreciated. I am using iPlanet App Builder 6.0 with iAS
6.0 SP1.
Thanks
Bhaskara Murthy
Eclipse Networks Inc

"Bhamu" <[email protected]> wrote in message
news:9keuo4$[email protected]..
I am trying to develop an entity bean attached to a key which have a
composite key using Primary Key Class. When I use the findByPrimaryKey
method of the bean it throws an exception as follows,I notice that you are using the reference CMP plugin.
at com.netscape.server.ejb.SQLPersistenceManager.find(Unknown Source)I'm also willing to bet that you may have created your own primary key class
with a single field, rather than using the fields type as a primitive
primary key class. If you do this, then SQLPersistenceManager will break.
It is badly written, and has some stupid assumptions. The one I remember
the most is that if you only have one primary key field, then
SQLPersistenceManager assumes you have used a primitive type such as
java.lang.String or java.lang.Integer, to represent it, rather than creating
your own <Enity>Pk.java file.
SQLPersistenceManager works for toy examples, but in general I would say it
is broken and unusable. Either use BMP, or splash out the money for Coco
Base from Thought Inc. Currently the only CMP plugin for iPlanet App server,
other than the reference implementation, that I know of.

Violation of Primary Key after a successful Update of BP

When I've updated a BP via VB.net is there something I should do with the BP object after I've done an update?
I can't seem to do consecative updates or I get the following error:
Violation of PRIMARY KEY constraint 'CRD1_PRIMARY'. Cannot insert duplicate key in object 'CRD1'
This is the code I use for updating:
For i = 0 To oBusinessPartner.Addresses.Count - 1
            oBusinessPartner.Addresses.SetCurrentLine(i)
            If (oBusinessPartner.Addresses.AddressType = SAPbobsCOM.BoAddressType.bo_ShipTo) Then
                '** Check only SHIP TO Addresses
                If (oBusinessPartner.Addresses.AddressName = cbSAPShipTo.SelectedItem) Then
                    '** Address User wishes to update
                    oBusinessPartner.Addresses.AddressType = SAPbobsCOM.BoAddressType.bo_ShipTo
                    oBusinessPartner.Addresses.AddressName = lSAPAddressItems(iIndex).Company
                    oBusinessPartner.Addresses.Street = lSAPAddressItems(iIndex).Address1
                    oBusinessPartner.Addresses.City = lSAPAddressItems(iIndex).City
                    oBusinessPartner.Addresses.ZipCode = lSAPAddressItems(iIndex).Postal
                    oBusinessPartner.Addresses.State = lSAPAddressItems(iIndex).Prov
                    oBusinessPartner.Addresses.Country = lSAPAddressItems(iIndex).Country
                End If
            End If
        Next
        lRetCode = oBusinessPartner.Update()
I check to make sure the address i want to up date is a ShipTo address, then I check to make sure that the item I am editing matches the one in the BPObject's CurrentLine.
Then I just change the values and .Update() it
After I've done this though, if I try to update another ShipTo address in the same BP, it comes up with that error.
Should I kill the BP object and reinitialize it?

Hi Tyler,
I totally understand what you wanna do.
The clauses which you use is trying to add a new address in the name of existing address.
You have to do 'setcurrentline' clause, I guess.
But now, you don't know 'what is the ShipTo line'.
Then, I think you'd better use some simple, easy but timely expensive way.
1. Count the records in CRD1 under conditions of the cardcode.
-> then you can get the number of addresses of the business partner.
2. use a loop (while, for.. doesn't matter.)
-> Now you can use setcurrentline clause.
3. If the addressname is 'ShipTo', then you can update the record, and exit loop.
pseudo code comes below.
sql = "Select count(*) FROM OCRD WHERE cardcode = 'aaa'"
ors.doquery sql
for i=1 to ors.fields.item(0)
oBusinessPartners.Addresses.setcurrentline(i)
if obusinessPartners.Addresses.AddressName = "ShipTo" then
oBusinessPartners.block = "newBlock"
exit for
end if
next i
Hope this useful for you.
Regards,
Hyunil choi.

Automatic Creation of primary key at the time of creating entries

Hi All,
I have a requirement to generate the primary key automatically at the time of creating entries. I have managed to do so by writing an event code for the maintainance object in SE54.
However called only when I create entries throught the Maintainance View.
It does not get called when I write an insert statement in the program.
Please suggest a way to achieve this.
Thanks in advance,
Archana.

for assigning the primary key while using insert statement, write a piece of code to get the last pk value, increment it by 1 and then assign this value to the pk field of the next record while inserting. if you are using a fn mod to insert values, just have this piece of code right before the insert query. a better way would be to use a GUID which will be unique globally. there is a fn mod GUID_CREATE which you can use to generate GUID. i don't think there is a way to automate this while making entries using se11, unless you change the code that actually creates the entry in the database.

Primary Keys and VARCHAR2 vs CHAR datatypes???

Could someone please tell me if there is documentation concerning the performance implications of creating a Primary Key using a VARCHAR2 datatype vs a CHAR datatype?

Well with a CHAR datatype, a value of ‘abc’ equates to ‘abc ’.
So that if the value of ‘abc’ already existed in the PK in the table and someone tried to Insert another row with a value of ‘abc ‘, it would return a unique constraint violation (the rule from the user being that a value with or without trailing spaces is the same value).
But if the PK column is defined as VARCHAR2, both the 'abc' and the 'abc ' could be Inserted into the Table because in the nonpadded comparison they are not equal.
Would defining the column as a VARCHAR2 with a Check Constraint to RTRIM the column be a better way to go?

Custom domains and primary keys

hi to all;
I've in troubles, with custom domains and the fact that sometimes, they must be as primary keys.
Searching in this news i see that we have to code, equals() and hashcode() methods, trying it i've got better behaviors , but still in problems.
My domain is an String , in "show time" i need it masked and query time ( as in database it is stored without mask ) i need it without any mask;
findByKey method i never know when it will work and when it will not;
Could some one point me to a right solution ?
fowards thanks

Dan:
Yes. Provide the correct equals() and hash() is the most important thing you need to do. There are other things you need to think about if the domain is NOT a scalar domain (single value). But, BC4J assumes that the PK is a scalar value, so your custom domain should be a scalar value value ==> equals and hash are two thing you'd need.
Thanks.
Sung

What index is suitable for a table with no unique columns and no primary key

alpha
beta
gamma
col1
col2
col3
100
1
-1
a
b
c
100
1
-2
d
e
f
101
1
-2
t
t
y
102
2
1
j
k
l
Sample data above and below is the dataype for each one of them
alpha datatype- string
beta datatype-integer
gamma datatype-integer
col1,col2,col3 are all string datatypes.
Note:columns are not unique and we would be using alpha,beta,gamma to uniquely identify a record .Now as you see my sample data this is in a table which doesnt have index .I would like to have a index created covering these columns (alpha,beta,gamma) .I
beleive that creating clustered index having covering columns will be better.
What would you recommend the index type should be here in this case.Say data volume is 1 milion records and we always use the alpha,beta,gamma columns when we filiter or query records
what index is suitable for a table with no unique columns and primary key?
col1
col2
col3
Mudassar

Many thanks for your explanation .
When I tried querying using the below query on my heap table the sql server suggested to create NON CLUSTERED INDEX INCLUDING columns ,[beta],[gamma] ,[col1]
,[col2] ,[col3]
SELECT [alpha]
,[beta]
,[gamma]
,[col1]
,[col2]
,[col3]
FROM [TEST].[dbo].[Test]
where [alpha]='10100'
My question is why it didn't suggest Clustered INDEX and chose NON clustered index ?
Mudassar

Key in View for two primary keys tables

Hi
I have some view joining LIKP and LIPS:
LIKP -   Primary Key MANDT, VBELN
LIPS -   Primary Key MANDT, VBELN, POSR
The Key for the LIKP-LIPS View is:
MANDT      LIKP
VBELN       LIKP
MANDT_I    LIPS
VBELN_I    LIPS
POSNR_I   LIPS
So, for that Key, if i need to acces a specify LIPS detail record , must i fill all the keys in the where (filling VBELN twice) like this:
WHERE VBELN = vbvalue AND VBELN_I = vbvalue AND POSNR_I = posvalue ?
How must i make the Select for acces this View using its Key ?
Regards
Frank

Hi
Just as I said:
SELECT SINGLE * FROM Z<VIEW> WHERE VBELN = <VBELN>
                                                            AND POSNR_I = <POSNR>.
Anyway it's should be better you correct the view deleting the field MANDT_I, VBELN_I, if you can't, I believe the select above works fine.
VIEW FIELDS:
View Field Table Field Key
MANDT LIKP MANDT Yes
VBELN LIKP VBELN Yes
<b>MANDT_I LIPS MANDT Yes <----- WRONG
VBELN_I LIPS VBELN Yes      <----- WRONG</b>
POSNR_I LIPS POSNR Yes
ERNAM LIKP ERNAM
ERDAT LIKP ERDAT
VKORG LIKP VKORG
PSTYV_I LIPS PSTYV
MATNR_I LIPS MATNR
WERKS_I LIPS WERKS
LGORT_I LIPS LGORT
By this defintion the field MANDT and VBELN are only twice, but it doesn't mean the record is double in the VIEW.
So it you have the delivery 100 with 3 items in the client 200, in the view you'll see:
Record 1: MANDT   = 200
               VBELN    = 0000000100
               MANDT_I = 200
               VBELN_I = 0000000100
               POSNR_I = 000010
Record 2: MANDT   = 200
               VBELN    = 0000000100
               MANDT_I = 200
               VBELN_I = 0000000100
               POSNR_I = 000020
Record 3: MANDT   = 200
               VBELN    = 0000000100
               MANDT_I = 200
               VBELN_I = 0000000100
               POSNR_I = 000030
Max

Better Primary Key

Similar Messages

Maybe you are looking for