Cross-server SQL

问题:

I want to port data from one server's database to another server's database. The databases are both on a different mssql 2005 server. Replication is probably not an option since the destination database is generated from scratch on a [time interval] basis.

Preferebly I would do something like

insert *
from db1/table1
into db2/table2
where rule1 = true

It's obvious that connection credentials would go in somehwere in this script.

回答1:

I think what you want to do is create a linked server as per this msdn article. You would then select using a 4 part object name eg:

Select * From ServerName.DbName.SchemaName.TableName

回答2:

You can use Open Data Source Like this :

EXEC sp_configure 'show advanced options', 1
GO
RECONFIGURE
GO

EXEC sp_configure 'Ad Hoc Distributed Queries', 1
GO
RECONFIGURE
GO


SELECT  *
FROM    OPENDATASOURCE('SQLOLEDB',
                   'Data Source=<Ip Of Your Server>;
                    User ID=<SQL User Name>;Password=<SQL password>').<DataBase name>.<SchemaName>.<Table Or View Name>

Go

回答3:

Are SQL Server Integration Services (SSIS) an option? If so, I'd use that.

回答4:

Would you be transferring the whole content of the database from one server to another or just some data from a couple of tables?

For both options SSIS would do the job especially if you are planning to to the transfer on a regular basis.

If you simply want to copy some data from 1 or 2 tables and prefer to do it using TSQL in SQL Management Studio then you can use linked server as suggested by pelser

1. Set up the source database server as a linked server
2. Use the following syntax to access data

select columnName1, columnName2, etc from serverName.databaseName.schemaName.tableName

回答5:

Well I don't agree with your comment on replication. You can start a replication by creating a database from scratch, and you can control either the updates will be done by updating the available client database or simply recreating the database.

Automated replication will ease your work by automatically managing keys and relations.

I think the easiest thing to do is to start a snapshot replication through MSSQL Server Studio, get the T-SQL corresponding scripts (ie the corresponding T-SQL instructions for both publication and subscriptions), and record these scripts as part of a job in the Jobs list of the SQL Agent or as a replication job in the replications folder.

回答6:

You could go the linked server route.

you just can't use the select * into you have to do an insert into select.

I would avoid replication if you don't have experience with it as it can be difficult to fix if it breaks and can be prone to other problems if not properly managed.

Keep it simple especially if the databases are small.

回答7:

Can you use the Data Transformation Services to do the job? This provides all sorts of bolt-together tools for doing this kind of thing.

You can download the SQL Server 2005 feature pack from Microsoft's website here

回答8:

CREATE VIEW newR1 
AS
SELECT * from OPENQUERY ([INSTANCE_NAME], 'select * from DbName.SchemaName.TableName')