| Author |
Topic |
|
jmh9403
Starting Member
6 Posts |
 Posted - 2014-07-01 : 19:51:53
|
I am setting up 3 servers for P2P replication
CMC-A server 1
CMC-B server 2
CMC-C server 3
All are running windows 2008R2
All are running SQL 2012 Inter prize
Server 1 has the primary database and tables
I know I have to set on each table the IDENTITY (1,1) key Col on the servers have to be different
The primary server 1 is being used all the time
The problem I have is setting up P2P replication to servers 2 and 3
Replication requires I make a backup of server 1 and restore it to 2 and 3.
Now all IDENTITY Col are the same
How can I change the IDENTITY Col for each server
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
jmh9403
Starting Member
6 Posts |
Posted - 2014-07-02 : 14:36:59
|
I checked it out and understand how it increments the IDENTITY col
Where do I put the code for each table
Would I put the code in the stored procedures for Subscribers
I'm kind of lost now
|
 |
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
Posted - 2014-07-02 : 14:40:52
|
You would just run it one-time for each replicated table on each server.
Tara Kizer
SQL Server MVP since 2007
http://weblogs.sqlteam.com/tarad/ |
|
|
|
jmh9403
Starting Member
6 Posts |
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
jmh9403
Starting Member
6 Posts |
Posted - 2014-07-03 : 17:01:16
|
When I use DBCC CHECKIDENT my table of 500 rows
I give it 501 - 2000 ID using DBCC CHECKIDENT
If the table goes up to lest say 800 rows
now I have used 300 assigned ids
If it goes back to 500 rows do I get back the 300 ids
that I assigned are does it work that way
|
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
Posted - 2014-07-05 : 16:50:43
|
No, it doesn't work that way. It is always increasing (or decreasing if you set it up that way, very rare). You need to account for this by setting the appropriate ranges for each server and possibly using the bigint data type.
Tara Kizer
SQL Server MVP since 2007
http://weblogs.sqlteam.com/tarad/ |
|
|
|
jmh9403
Starting Member
6 Posts |
Posted - 2014-07-13 : 13:22:50
|
I under stand using DBCC CHECKIDENT
I would like to get away from using it because of the limitation of running out of ID's
Could I Export the database from the primary server (which it's ID's are 1,1) to another database and it's ID's would be 1,3
Wants I get everyone on the new database. I would do the same thing for server 2 and the ID's for this server would be 2,3
and server 3 and it's ID's would be 3,3
Server 2 and 3 are not being used yet and have no data
If what I am trying to do is not practical I will use DBCC CHCKIDENT |
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
Posted - 2014-07-13 : 18:59:15
|
You will never run out of ids if you use bigint.
Tara Kizer
SQL Server MVP since 2007
http://weblogs.sqlteam.com/tarad/ |
|
|
|
jmh9403
Starting Member
6 Posts |
Posted - 2014-07-14 : 09:06:44
|
| I want to tell I really appreciate your help |
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
|
P2P replication
