Comments for sqlity.net

Comment on Presentation Materials – Index Internals by 2012-04-12: Index Internals – SQL PASS Regionalgruppe Hamburg - sqlity.net | sqlity.net

Mon, 09 Apr 2012 10:35:54 +0000

[...] TitleIndex Internals – How SQL Server stores and accesses data [...]

Comment on The Gap in the Identity Value Sequence by A mock too far? What is good practice for mocking database objects with tSQLt?

A mock too far? What is good practice for mocking database objects with tSQLt? — Wed, 28 Mar 2012 08:10:50 +0000

[...] been rolled back – potentially leaving gaps in the IDENTITY sequence. As Sebastian Meine says here, you cannot depend on unbroken identity sequences in a multi-user system (or when rollbacks may [...]

Comment on The Gap in the Identity Value Sequence by Unit Testing Databases with tSQLt Part 10 – testing a foreign key’s ON DELETE or ON UPDATE actions

Tue, 20 Mar 2012 20:21:51 +0000

[...] primary key of the parent table is also an IDENTITY column. As Sebastian Meine points out in “The Gap In The Identity Value Sequence“, identity values are only used once and, in a multi-user system, you should not rely on [...]

Comment on Presentation Materials – Index Internals by 2012-03-10: Index Internals – SQL Saturday #110, Tampa, Florida | sqlity.net

Tue, 13 Mar 2012 14:13:14 +0000

[...] TitleIndex Internals – How SQL Server stores and accesses data [...]

Comment on Purge Problems [TSQL Tuesday #027 - The Big Data Valentine’s Edition] by The T-SQL Tuesday #027 Roundup « Voice of the DBA

The T-SQL Tuesday #027 Roundup « Voice of the DBA — Tue, 21 Feb 2012 17:03:48 +0000

[...] Meine wrote an interesting piece on purging data, and how large data manipulation operations can benefit from [...]

Comment on Presentation Materials – Index Internals by dblumer

dblumer — Thu, 09 Feb 2012 23:26:31 +0000

Great presentation Sebastian. Thanks a lot. Too bad we didn’t have time to wrap up. I wanted to hear you talk about the Seek vs Scan slide. Hopefully we can catch up and discuss it a little bit more soon.

Comment on Selecting the entire Database as XML String – 2 by daniel

daniel — Mon, 06 Feb 2012 11:31:27 +0000

Hi Sebastian,

may I ask you to take a look at this: http://stackoverflow.com/questions/9159200/bulk-instert-into-ms-sql-2008r2-from-xml-raw-data

Thanks a lot for your help so far.

Comment on Selecting the entire Database as XML String – 2 by daniel

daniel — Mon, 06 Feb 2012 08:45:42 +0000

Thanks Sebastian, this really works. And I consider your opinion on schemas valuable, so we will add them to the query as in your original solution.

Comment on Selecting the entire Database as XML String – 2 by Sebastian

Sebastian — Fri, 03 Feb 2012 14:36:20 +0000

Yes, you are right. If you want all columns to be attributes (like here) you can use FOR XML RAW. I usually don’t think of it because of it’s restricted flexibility.

To remove the tag, just replace the line containing the SET statement in your query with this one:
SET @cmd = 'SELECT [@name],(SELECT data) FROM('

On a side note: In SQL Server you should always specify the schema name of all objects. Not doing so can have a lot of unwanted consequences, ranging from performance impacts to hard to find bugs. So if all your tables are in the “dbo” schema you should at least hardcode it in your query, but then you might as well use the OBJECT_SCHEMA_NAME function to be prepared for future additions to your database.

Comment on Selecting the entire Database as XML String – 2 by daniel

daniel — Thu, 02 Feb 2012 08:01:15 +0000

Hi Sebastian, I think there might be an easier approach as opposed to getting all the column names into the dynamically constructed query by using for XML RAW. I modified your original SQL query like below (I left out the quotename and schema operations as I do not need them in my implementation).


DECLARE @cmd NVARCHAR(MAX)

SET @cmd = 'SELECT * FROM('
    + STUFF(
(SELECT ' UNION ALL SELECT ''' + name + ''' [@name],' + '(SELECT * FROM '
        + name
        + ' FOR XML RAW,BINARY BASE64,TYPE) data'
 FROM sys.tables
 FOR XML PATH,TYPE).value('.', 'NVARCHAR(MAX)'), 1, 11, '')
    + ')X FOR XML PATH(''table''),TYPE;'
 
EXEC(@cmd)

The only problem is that there's the data node inside all table nodes that I cannot rid of (yeah, my SQL skills are not up to par with yours...)