Comments on: SQL SERVER – Discussion – Effect of Missing Identity on System – Real World Scenario

By: SQL SERVER – Understanding Identity Beyond its Every Increasing Nature – Quiz – Puzzle – 3 of 31 « SQL Server Journey with SQL Authority

Wed, 04 Jan 2012 01:32:52 +0000

[...] of Record Reseed Identity of Table – Table Missing Identity Values – Gap in Identity Column Discussion – Effect of Missing Identity on System – Real World Scenario Negative Identity Seed Value and Negative Increment Interval Enable Identity Insert – Import [...]

By: Jacob Sebastian

Jacob Sebastian — Tue, 18 Aug 2009 02:28:07 +0000

Most applications will obtain a valid key from the table (either by a query or calling SCOPE_IDENTITY() after a new insert) before performing an insert operation to a referenced table. This makes sure that only valid references are inserted to the referenced table.

the question being discussed here is the implications of having non-sequential IDENTITY values. What does it matter if the customer table has identity values ’1, 2, 3′ and ’5′. (missing 4). In which real-life scenario some one will assume that customer ’4′ exists and try to perform an operation?

By: Jay

Jay — Mon, 17 Aug 2009 17:41:43 +0000

Hi all,We rececntly encountered the problem in one of the application.
We have a table with identity column. Its referenced by a foreignkey constraint from another table.
Table1(Id,event,time) ID is identity.
Table2(Id,table1_ID,message). Id is identity. talbe1_ID is the foreignkey that references Id from table1.
Our application was getting error messages(below) when they were trying to insert a particular value in table2 because the value was missing in table1.
Error 547, Level 16, State 0, Procedure storedprocedure, Line x, Message: The INSERT statement conflicted with the FOREIGN KEY constraint “FK_table1_ID”. The conflict occurred in database “xxx”, table “dbo.table1″, column ‘Id’..

Later we found out some identity values were missing in the primary tables due to some transactions that were rolled back.
Thanks
Jay

By: Jacob Sebastian

Jacob Sebastian — Thu, 13 Aug 2009 13:46:14 +0000

I agree with Ron. Any data migration project should maintain the same Primary Key values. In a serious application there may be hundreds of tables where an ID is linked and if you re-sequence it, the whole data will be a mess.

By: Ron

Ron — Tue, 11 Aug 2009 10:28:25 +0000

Gangadhar,

When migrating data with identity values use SET IDENTITY_INSERT IdentityTable ON.

By: Gangadhar Naidu

Gangadhar Naidu — Tue, 11 Aug 2009 08:39:44 +0000

Yes I know a real-world scenario where a Missing Identity value in any table can create problems.

At the time of data migration from one server to another server.

see the example:
server1:
table1(ID,Name) where ID is identity
table2(ID,MultipleAddress) where ID is foreignkey from table1.

example Data:
table1:
1, Gangadhar
3, Imran
4, pinaldave

table2:
1, Address1
1, Address2
3, Address1
3, Address2
3, Address3
4, Address1

Now i started migration from server1 to server2
migrated: server1.table1 to server2.table1
The data migrated like below:
1, Gangadhar
2, Imran
3, pinaldave

migrated: server1.table2 to server2.table2
table2:
1, Address1
1, Address2
3, Address1
3, Address2
3, Address3
4, Address1

Now how do you get address for relevent ID.

By: Kashif Iqbal

Kashif Iqbal — Tue, 11 Aug 2009 04:40:28 +0000

Nice information. I am a student of Software Engineering. I am very much interested in search papers, but i have no one to guid me. will you help me that how can i start?