Compare data in Tables

I am trying to determine the changes an application makes to a database.
The plan is to copy the existing schema (active) to a reference schema, run
the application and then diff the table data between the reference and the
a active schema. I have found one software vendor who has a tool to do
this, but it will only do one table at a time (interactively); I have more
then 300 and will run this a few times.

One other way of determining the changes, I guess, would be to log all sql
statements (in order), but I don't know how to do this (either).

Any pointers would be greatly appreciated.

LeoIf its changes you are looking for try running SQL Profiler against it.
Filter for where writes > 0. Another solution would to write a script to
doing the all tables comparison. Something like

Create a table with tablename, checksumbefore, checksumafter, rowsbefore,
rowsafter, numberofrowsdiff
Write a cursor of all user tables
For each table
Get count of rows with select count(*)
calc the CHECKSUM of each row and write to individual temp tables
select count(*) from checksumafter where checksum not in checksumbefore
insert/update the table

By the end of the script you should have indentified which tables change and
by how much.

"Leo" <leolist@.optushome.com.au> wrote in message
news:Xns9639579AFB3Bleolistoptushomecoma@.211.29.13 3.50...
>I am trying to determine the changes an application makes to a database.
> The plan is to copy the existing schema (active) to a reference schema,
> run
> the application and then diff the table data between the reference and the
> a active schema. I have found one software vendor who has a tool to do
> this, but it will only do one table at a time (interactively); I have more
> then 300 and will run this a few times.
> One other way of determining the changes, I guess, would be to log all sql
> statements (in order), but I don't know how to do this (either).
> Any pointers would be greatly appreciated.
> Leo|||There is a software tool that can do this for you called DB Ghost
(www.dbghost.com). Its very fast at comparing data and can be run from
the command line for a fully automated process. A single command will
do any number of tables that you desire.

It's also the cornerstone of a full change management solution for SQL
Server databases i.e. it can build, compare and synchronize the schema
AND data directly from drop/create scripts held in a source control
system.

I highly recommend you check it out.|||"Malcolm" <malcolm.leach@.innovartis.co.uk> wrote in
news:1113551521.457015.133450@.l41g2000cwc.googlegr oups.com:

> DB Ghost

DB Ghost did exactly what I needed.

Thanks for your advice

Leo

Command Line Functionality

Is it possible to parse and/or find out the estimated execution plan of an SQL Query at command line without making use of the IDE ?

See BOL. Search the index for SET SHOWPLAN. You will find various options.

See SQLCMD for command line execution.

Combine the two and you're golden.

comma delimited list

Hello,
I have created a small query that will output a list of email addresses that
are separated by commas. I plan to copy and paste the list into a email
invite field (send to: field for an email app). But what is happening is tha
t
only one address appears in the field when I paste because of the added
spaces between the names, example currently outputs like:
blahblah@.blah.com,
blahblah@.blah.com,
blahblah@.blah.com
What I need is it to output like this:
blahblah@.blah.com,blahblah@.blah.com,blahblah@.blah.com
Here is the small query that I am using to extract the emails.
SELECT RTRIM(LTRIM(EMAIL_ADDRESS)) + ',' AS [Name]
FROM evPRUclinicala
WHERE CREATE_DATE BETWEEN '4/1/06' AND '4/30/06'
AND EMAIL_ADDRESS IS NOT NULL
Any help would be appreciated and thanks in advance.
Message posted via http://www.webservertalk.comIn general, a recommended approach is to extract the resultset outside the
server and massage the data to appropriate display format using some client.
Regarding the alterantives for doing it at the server, you can check out the
following links:
( For SQL 2005 only )
http://groups.google.com/group/micr...br />
9b9b968a
( For SQL 2000 & 2005 )
http://groups.google.com/group/micr...br />
6dd9e73e
Anith|||Anith Sen wrote:
>In general, a recommended approach is to extract the resultset outside the
>server and massage the data to appropriate display format using some client
.
>Regarding the alterantives for doing it at the server, you can check out th
e
>following links:
>( For SQL 2005 only )
>http://groups.google.com/group/micr...r />
a9b9b968a
>( For SQL 2000 & 2005 )
>http://groups.google.com/group/micr...r />
66dd9e73e
>
Thanks for the direction, but I am still have some problems. I elected to us
e
the cursor version of your solution: see below
DECLARE @.temptable TABLE ( list VARCHAR( 8000 ) )
SET NOCOUNT ON
DECLARE @.PRUemail VARCHAR( 1000 ), @.emailNext VARCHAR (100)
DECLARE c CURSOR FOR
SELECT EMAIL_ADDRESS
FROM evPRUclinicala
ORDER BY EMAIL_ADDRESS
OPEN c
FETCH NEXT FROM c INTO @.emailNext
SET @.PRUemail = @.emailNext
WHILE @.@.FETCH_STATUS = 0 BEGIN
IF @.emailNext > @.PRUemail BEGIN
INSERT @.temptable SELECT @.PRUemail
SELECT @.PRUemail = @.emailNext
END ELSE
SET @.PRUemail = COALESCE( @.PRUemail + ',', SPACE( 0 ) ) +
@.emailNext
FETCH NEXT FROM c INTO @.emailNext
END
INSERT @.temptable SELECT @.PRUemail
CLOSE c
DEALLOCATE c
SET NOCOUNT OFF
SELECT * FROM @.temptable
But I am still getting a vertical list with no commas. Can you tell me what
I
am doing wrong? Thanks
Message posted via http://www.webservertalk.com|||>> I elected to use the cursor version of your solution:
Why? It is a bad choice to begin with.
I have no idea how the table evPRUclinicala is structured or what the nature
of the data is. Can you post your table DDLs & a few sample data? For
details, refer to: www.aspfaq.com/5006
Anith|||"Jay via webservertalk.com" <u7124@.uwe> wrote in message
news:601bd1f4c21b5@.uwe...
>.
> Can you tell me what I am doing wrong?
>.
Yeah, your not using Rac :)
www.rac4sql.net/onlinehelp.asp?topic=236|||Anith Sen wrote:
>Why? It is a bad choice to begin with.
>
>I have no idea how the table evPRUclinicala is structured or what the natur
e
>of the data is. Can you post your table DDLs & a few sample data? For
>details, refer to: www.aspfaq.com/5006
>
What would have been a better choice? If I should use a different method tha
n
I would appreciate your advise here. If not I will get you the DDLs.
Message posted via webservertalk.com
http://www.webservertalk.com/Uwe/Forum...amming/200605/1|||Cursors are generally performance hogs. In most cases, it is wiser to
utilize a set based solution in SQL where a set of rows are manipulated as a
whole rather than doing it one row at a time.
As a long term solution, consider returning the resultset to the client and
manipulate the data using a client application program -- this allows the
stored procedure to be sufficiently generic, eliminates the need to use
server's processing power for string manipulations & loop-based constructs
and leverages the string functionalities of the client language, which in
most cases are more versatile than t-SQL.
If you are convinced to use the server for such stuff, consider the table
valued UDF approach which might be better performing for medium sized
datasets. For SQL 2005, the XML kludge might be faster, though we do not
have much useful benchmarks with this approach.
Alternatively, for large volume transactions of this nature, consider using
3rd party tool's like RAC mentioned in Steve's post.
Anith|||Anith Sen wrote:
>Why? It is a bad choice to begin with.
>
>I have no idea how the table evPRUclinicala is structured or what the natur
e
>of the data is. Can you post your table DDLs & a few sample data? For
>details, refer to: www.aspfaq.com/5006
>
Ok I will look at the other solutions rather than the one chosen. Thanks
Message posted via webservertalk.com
http://www.webservertalk.com/Uwe/Forum...amming/200605/1