Is this good approach for generating Primary Keys ?

Hi,
All our EJBs are state less Session Beans.
We have created a state less Session(name SequenceEJB) for getting Primary Key.
Thre is a method,
int getSequenceNumber(String tableName,String fieldName) in SequenceEJB. And the following query is used in the above said method.
SELECT MAX("+fieldName+") ID FROM "+tableName
Each EJB will lookup SequenceEJB, and call the above said method for getting sequence number.
Is this approach sclable ?
Can we proceed by this approach ?
Thanks in advance
Srinivas

As long as you are in EJB, synchronization won't help you because it is not permitted by the EJB spec.
If you can go with JDBC 3.0 / Java 1.4, the statement interface supportes generated keys. See http://java.sun.com/j2se/1.4/docs/api/java/sql/Statement.html.
If you cannot go with JDBC 3.0 / Java 1.4, you can implement the key-generation db-specific (e.g. using oracle sequences for oracle tables, and whatever exists for DB2, SQL Server etc.) That of course requires special support for all databases, but at least you can run your app with different databases.
Or you can try to do it the following way, using a separate table for the keys.
ResultSet rs = statement.executeQuery("select id from mySequences");
oldId = rs.getInt(1);
newId = oldId + 1
int updatedRows = statement.executeQuery("update mySequences set id = newId where id = oldId");
if(updatedRows == 1) {
// succesful
} else {
// not sucessful, somebody else has already retrieve this id; do something like Thread.sleep() and retry
}This way, the database does the actual synchronization between the processes (each process sees the same data in the tables), and you implement the retry logic.

Similar Messages

How do I obtain the next number for a Primary Key using an ADF View Object?

I have two separate View Objects (A & B) for the same Entity Object. View Object A does a SELECT on all of the fields in the table. This View Object is where I execute my adds and updates. View Object B is only used to retrieve the next number for the primary key. This is done so that when I add a row to the database, I always get the max number of the primary key and add one to it. I accomplished this by setting the SQL mode to Expert and using the SQL: "SELECT MAX(NBR) AS MAX_NUMBER FROM TABLE_1". This may be overkill having a seperate View Object for this, but so far this is the only way I have found to obtain the next number. However, I have discovered that this way does not always work.
The problem I'm running into is when I try to add multiple records to View Object A without committing the transaction between each add. Because View Object B is disconnected from View Object A, the MAX_NUMBER of View Object B comes back with the same number for each add I do on View Object A. So I know I must retrieve the MAX_NUMBER from View Object A.
I've tried using the following code in my Table1ViewImpl class:
this.setQuery("SELECT MAX(Table1.NBR) AS MAX_NUMBER FROM TABLE_1 Table1");
this.executeQuery();
The view object now has what I want, but I have yet to figure a way to extract the MAX_NUMBER out of the View Object. I've also looked into using the method addDynamicAttribute() but I can't figure out any way to set the attribute with the MAX_NUMBER.
I can't be the only one trying to retrieve the next number from a database table using ADF. Can anyone help me with this? FYI - I'm using JDev 10.1.3 EA.

You missing the point.
On a multi-user db knowing the next highest number doesn't guarantee the number will be available when it comes time to commit the record. You can prove this to yourself by opening two instances of your app and do whatever you do to add a new record to your VO. Both will assume the same number, and when you commit an error will be generated
You must use sequences to avoid the possibility of duplicate keys. If you are trying to avoid gaps in your numbering then you need to convince yourself why this is necessary.

Sequences to generate primary keys

Hi,
A question on using sequences to generate primary keys. We have a table and in all environments, test, uat, prod, a sequence exists to generate the primary key for the table. This table is a reference table and is not updateable by the users.
When developing scripts to insert records into the table, the developers are currently not using the sequence but typing in unique numbers so data scripts consist of something like this:
INSERT INTO EMPLOYEE_GROUP
(employee_group_id, code, description,...)
VALUES
(45,'sdf','sdfsdfsf');
INSERT INTO EMPLOYEE_GROUP
(employee_group_id, code, description,...)
VALUES
(46,'sdf','sdfsdfsf',...);
a) Is this normal practice?
b) Sometimes they are using the sequence to generate the primary key in development but then hardcoding the number in the script so in development the sequence might be at 160, but in the other environments the sequences will still be at one. Is this also normal practice?
My usual method in developing scripts is to use <seqeunce>.nextval but it's somewhat incompatitble in the current environment.

You are using a surrogate, or synthetic primary key. For a primary key generated in this way, the value is inconsequential. All that matters is that it is unique and the tables with foreign keys into it, if any, also have that value set properly. But it doesn't matter if the value is 1, 100, 12344, 2222222222, etc. All you care about is the uniqueness.
If you create the data properly then child records and their parents should be correct in all systems you mention, it's just that the actual value of the key will differ and that's fine. You care about the relationship of the records, not the value of the synthetic keys.

Checking for duplicate primary keys on row inserts

Checking for duplicate primary keys on row inserts
I have a situation where I will be making bulk table inserts knowing that the primary key value will in some cases already exist. In this is the case I simply want to ignore the duplicate inserts.
Should I be performing a sub-query on the table and using a statement like:
where not exist in
Or is there a cleaner way of discarding or checking for duplicates on insert.
My concerns were mainly one of performance, as my routine will be inserting a few thousand rows in its operation.

The MERGE commnad is a good option when a large percentage of the data will exist in the target because it is much more efficient to attempt to update then insert when the update affects zero rows than capture an error and convert it to an update.
However, since in this case it would appear that only a few rows will alreadys exist and we want to ignore the duplicates when they exist then
begin
insert
exception
when dup_value_in_index then null;
end
would be the way to code this one. The bulk insert version has in 9.2 the ability to store the errors so that they can all be handled at once which means the rest of the array insert can work.
HTH -- Mark D Powell --

Master-detail with auto-generated primary key(master)

Hello,
I have the following master-detail setup:
Master - block A primary key is: codeA
Detail - block B
codeA is genrated only at commit time and is obtained from a sequence.
User needs to add detail records. This is easy to do since he has to click on the insert icon on the toolbar. At commit time, use commit_form.
Prob: the requirements as asked by the client is to add a button (which we will label INSERT DETAIL). When the user click on the button, it opens a window where he can enter the detail info then click on save. This should bring him back to the master-detail form and the record in the detail block is inserted.
What is the problem here ? since the detail block has the foreign key tied to the master primary key, the window-form (with the call_form) that we opened for the detail entry won't be able to save since at this point we still have not assigned a value for the primary key (and hence a value for the foreign key).
Possible Suggestions:
-Use a temp database table to hold the detail record from the second form and use it to transfer values to the detail record in the master-detail
-Use a global record group
-We can't use parameter list to pass back these values.
So my question is:
What would be the most efficient way to have an insert button on the detail block that would allow the user to have a pop up where he can insert his values and then be brought back to the master-detail.
Thanks.

Hello again,
May be I was not clear enough.
Scenario 1: We use the master-detail form as is with the default oolbar. In this case, the user can insert the detail records one by one without needing the primary-foreign key value since this is handled by default.
Once we save the form (commit_form), I use the pre-insert trigger to get the master block primary key generated from the sequence and since the detail block key is copied from this value, both are saved correctly and it is the end of the story.
Scenario 2: As explained in the initial post, the user will populate the detail records one by one by clicking on the -INSERT DETAIL- button and hence has a window where he can insert the detail info and then be brought back to the master-detail form every time he enters a new detail record.
The problem here is that I can't generate the primary key for the master block since the client has the following requirement:
The user can always change his mind and not complete, meaning save the form, his process
As such, the key should be generated in the last step before Commit.

Creating a script for a PRIMARY KEY USING INDEX SORT doesn't work

Probably a bug.
h1. Environment
Application: Oracle SQL Developer Data Modeler
Version: 3.0.0.655
h1. Test Case:
1. Create a new table TRANSACTIONS with some columns.
2. Mark one of numeric columns as the primary key - PK_TRANSACTIONS.
3. Go to Physical Models and create new Oracle Database 11g.
4. Go to Physical Models -> Oracle Database 11g -> Tables -> TRANSACTIONS -> Primary Keys -> PK_TRANSACTIONS -> Properties:
a) on General tab set Using Index to BY INDEX NAME
b) on Using Index tab choose a tablespace
c) on Using Index tab set Index Sort to SORTED.
5. Export the schema to DDL script. For the primary key you will get something like this:
ALTER TABLE TRANSACTION
ADD CONSTRAINT PK_TRANSACTION PRIMARY KEY ( TRAN_ID ) DEFERRABLE
USING INDEX
PCTFREE 10
MAXTRANS 255
TABLESPACE TBSPC_INDX
LOGGING
STORAGE (
INITIAL 65536
NEXT 1048576
PCTINCREASE 0
MINEXTENTS 1
MAXEXTENTS 2147483645
FREELISTS 1
FREELIST GROUPS 1
BUFFER_POOL DEFAULT
) SORTED
h1. Reason of failure
The script will fail because SORTED is not allowed here. It should be SORT.
Additionally, the default behaviour for Data Modeler is to set Index Sort to NO but default setting for Oracle database 11g is SORT. Shouldn't Data Modeler use SORT as the default value?
Edited by: user7420841 on 2011-05-07 03:15

Hi,
Thanks for reporting this problem. As you say, it should be SORT rather than SORTED. I have logged a bug on this.
I also agree that, for consistency with the database default, it would be better to have SORT as the default in Data Modeler.
David

[Incoming Payments for WTax] Primary Key not exist in DB

While performing the TDS upgrade, I am getting an error message as "[Incoming Payments for WTax] Primary Key not exist in DB".
I have checked that the sheet conforms to the instructions provided in the Upgrade guide but am not able to resolve this.
Any ideas?
Regards,
Gyanesh

Hi,
It seems to be some field in the TDS upload Excel sheet which has primary key field not having the correct value.
Double check the same where Primary Key if linked with some value in other sheet, enter the same.
Kind Regards,
Jitin
SAP Business One Forum Team

Approach for generating a report while having Func Specs of the Report

Hi Experts,
Please tell me what should be the approach for generating a report when having Func Specs of the Report.
Thnx
Sid

Hi,
1st you need to know the business need and related data.
1st know that according to the business need is there any standard SAP report is there or not.
try to know the Data base table and related fields required according to the need.
know the relations of the data base tables to map the data.
Accordingly prepare Technical spec's or go a head for coding.
Thanks.
If this helps you reward with points.

Duplicate entries for same primary key

Hi,
I am facing problem to insert 2 or more than 2 entries for same PRIMARY KEY in the database table.
As I know that we cant do that. But Client has given me the XL sheet which contains 2 entries for same primary key. How can it be done? Please let me know how can I insert 2 data for same primary key in database table.
Waiting for your answers.
Thanks in advance.
Regards,
Prasanna

Hi,
You can achieve this .... All you have to do is to add a new field (a Sequence Number ) to the table. This number will be incremented for each duplicate primary keys. For example....
Consider the excel file has duplicate entries 3 primary keys. Now you add a new field named Sequence Number in your DB Table.....
When you load the data into Database...then the records will look like this...
Key1     Key2    Key3    Seq
A          B          C           1
A          B          C           2
A          B          D           1
A          B          D           2
A          C          D           1
A          C          E           1
and so on...
I hope this solves your purpose.....
Whenever there are duplicate entries, such as the one mentioned in your scenario, then a new field can be added in Database. This field acts like a count or sequence number.... Thus you can maintain unique records.
Regards,
Vara
Regards,
Vara

How generate primary key atumatically in jdeveloper 10g

Hi,
I am using jdev 10g. I created a fusion web applicaton.
I am not getting any option to generate primary key from sequence automatically.
In Jdev11g we can create by trigger easily.
Please help
Thanks
Raj

Hi Raj,
Check this link
[http://download.oracle.com/docs/html/B25947_01/bcentities006.htm#sm0147|http://download.oracle.com/docs/html/B25947_01/bcentities006.htm#sm0147]
Regards,
Vikram

Generate primary key automatically

Hi, I'd like to know how to generate primary keys automatically with JHeadStart because when I try to make an insert into a table, the page gives me the option of insert the primary key of the register.
Thanks a lot

I have a table which have 2 fields, one of those is the primary key, the other is the description.
I have established the primary key as a DBSequence, but when I try to insert a new record testing the service, appears a negative number in the primary key and instead of that I want to appear the correspondent key, or if doesn't appear the correct value, the insert be done correctly.
At least I change manually the value if this key, the insert can't be done.

Exception: non-read-only mapping defined for the primary key field

Hello.
I'm new to Oracle and I created Java EE Web Project.
And I created entities from tableA, tableB, tableC from my database.
tableA has foreign key to tableB and has unidirectional Many-to-One relationship. tableC has two primary keys and one of these is foreign key to primary key in tableA. tableC also has unidirection Many-to-One relationship to tableA.
Then I created session bean and data control from it, in design window, created read-only table from tableA.
And I selected columns in tableA and also in tableB.
I ran the application and saw the following exception in log window.
Local Exception Stack:
Exception [EclipseLink-46] (Eclipse Persistence Services - 2.1.3.v20110304-r9073): org.eclipse.persistence.exceptions.DescriptorException
Exception Description: There should be one non-read-only mapping defined for the primary key field [tableC.tableA_ID].
Descriptor: RelationalDescriptor(mypack.tableC --> [DatabaseTable(tableC)])
tableA_ID is a primary key in tableA.
How can I solve this error?
Please help me.
Edited by: 900471 on 2011. 12. 3 오전 5:32
Edited by: 900471 on 2011. 12. 3 오전 5:33
Edited by: 900471 on 2011. 12. 3 오전 5:33
Edited by: 900471 on 2011. 12. 3 오전 5:34

There are not enough details to be sure since you have not provided the mappings. From just the error, it looks like you are using the tableC.tableA_ID field as the foreign key in the ManyToOne relationship to A, but have marked it as insertable=false, writeable=false, meaning that it cannot be updated or used for inserts.
Either make it writable (set the settings to true), or add another basic mapping/attribute in the entity for TableC that maps to the field which you can use to set when you insert a new tableC entity. A simple example is available at
http://wiki.eclipse.org/EclipseLink/Examples/JPA/2.0/DerivedIdentifiers
Best Regards,
Chris

How to specify tablespace for a primary key inde in create table statement

How to specify the tablespace for a primary key index in a create table statement?
Does the following statement is right?
CREATE TABLE 'GPS'||TO_CHAR(SYSDATE+1,'YYYYMMDD')
            ("ID" NUMBER(10,0) NOT NULL ENABLE,
            "IP_ADDRESS" VARCHAR2(32 BYTE),
            "EQUIPMENT_ID" VARCHAR2(32 BYTE),
            "PACKET_DT" DATE,
            "PACKET" VARCHAR2(255 BYTE),
            "PACKET_FORMAT" VARCHAR2(32 BYTE),
            "SAVED_TIME" DATE DEFAULT CURRENT_TIMESTAMP,
             CONSTRAINT "UDP_LOG_PK" PRIMARY KEY ("ID") TABLESPACE "INDEX_DATA"
             TABLESPACE "SBM_DATA";   Thank you
Edited by: qkc on 09-Nov-2009 13:42

As orafad indicated, you'll have to use the USING INDEX clause from the documentation, i.e.
SQL> ed
Wrote file afiedt.buf
1 CREATE TABLE GPS
2              ("ID" NUMBER(10,0) NOT NULL ENABLE,
3              "IP_ADDRESS" VARCHAR2(32 BYTE),
4              "EQUIPMENT_ID" VARCHAR2(32 BYTE),
5              "PACKET_DT" DATE,
6              "PACKET" VARCHAR2(255 BYTE),
7              "PACKET_FORMAT" VARCHAR2(32 BYTE),
8              "SAVED_TIME" DATE DEFAULT CURRENT_TIMESTAMP,
9               CONSTRAINT "UDP_LOG_PK" PRIMARY KEY ("ID") USING INDEX TABLESP
ACE "USERS"
10               )
11*              TABLESPACE "USERS"
SQL> /
Table created.Justin

How to change the source type for a primary key on a form?

Hi,
At the time of creating a form, I had set the source type for the primary key to an existing sequence.
Now I want to change the source to a trigger.
Can anyone suggest how to do it?
Thanks in advance,
Annie

Annie:
Define the trigger and then delete the page process named 'Get PK'
Varad

How to pass values for compound primary keys for updating the record in db.

i have a entity bean in my ejb with two primary keys to it...eg. roleid,rolename...
when iam retriving the record i use only one primary key ...so again to modify and update he same record which i retrived is a pbm..
i cannot set the values for the compoundpk,,,
but i can do the update for the bean with single primary key..
i need to know how to pass the values/parameters for the coumpound primary key in my update method...
thanx in advance...
Arjun.G

I istantiated a object for the pk class and assign the values for the primary key fields...
Arjun.G

Is this good approach for generating Primary Keys ?

Similar Messages

Maybe you are looking for