I'm importing a set of tables from a SQL Server DB to Power BI Desktop.
It's a huge database and I'm selecting only some of its tables.
However, these tables have empty columns or columns with zeros only, so I've created two M functions to apply to each and every table to "clean" them.
Is it possible to specify an SQL command in the import settings to fetch only "valid" columns, thus avoiding loading them to Power BI and use custom functions?
Can anyone please share sucha a SQL command?
Thank you!
You can specify an initial SQL query in the connection object. If you click on the settings for it and open the advanced section, you can enter it there (screenshot is from the MySQL connector but I'm sure the SQL Server one is similar).
The output code would look something like
SQL.Database("sql.server.address:1234", "your_sql_server", [Query="SELECT columns, you, want FROM table_you_want"])
However note that if you are using the native connector in Power Query/BI, it attempts to convert whatever you do in Power Query into SQL on the back-end. If you right click on steps in your PQ process and click "Native Query" it will show you the SQL code it is converting to. If at some point that is greyed out it means whatever you were doing was something it couldn't convert.
I'm looking for a way to filter the data to employees something like create a "user to Client relationship", the problem is this is after the fact,I have 100+reports that already exist and I don't want to edit each one, I use a shard Data source, can i somehow implement RLS on that source ? or maybe use ssas RLS in conjunction.
SQL server 2012 enterprise,
Thank you
One solution which comes to my mind is to add to each shared dataset a part which will filter rows based on domain user name.
This will require that:
User running SSRS report will have to have access to SQL Server
You would have to build tables to define access right (if you do not have already)
If you would be running SQL Server 2014 and onwards you could use built in RLS. You would have to define access right and create access control function. Still you have to have some kind of data store which will help you to determine what is visible to each user - more details on RLS here.
Can I pinpoint SQL Profiler just to one Database, my server has around 50 Database and I just want SQL Profiler to check one Database because the site that uses that Database has gone very slow.
Yes, you can.
When you set your trace up, go to the Events Selection tab and click the "Show all columns" checkbox. Then the Database column appears and you can put a filter on the database name in Column Filters.
I want to save any kind of log/tables with every query executed by my application.
I know I could do this by coding it (before I make any query, I insert a new row in a log table with the query and the user who is executing it.
I have read it can be done automatically but I'm not sure how can it work with WCF Services. I mean every query is going to be executed by the same SQL user and this wouldn't be very useful for audit operations (I need to know WHO made every query, and users will be validated against my own users tables).
Have you ever had a similar scenario? Thanks in advance!
As a starting point it may be worth looking into doing this via SQL Server Profiler. You can normally find this in the Tools Menu in Management Studio.
You can set up a trace to capture all SQL run on a server. More importantly you have a myriad of filter options which can be applied so that you only capture the data you are interested in (e.g. DatabaseName, UserName).
This information can be stored directly in a SQL Table, which should give you the abillity to join onto. Of course running anything like this will result in some overhead on the SQL box.
You can try the SQL Server Audit feature. It audits singe or groups of events both on server and database level. However, be advised that the database level auditing is available in SQL Server Enterprise and Developer editions only
I have managed to get SQL Server 2005 Express up and running on my computer Ok in order to do some testing before trying this in the "Real World".
I have a fairly large MS Access 2007 Database application I need to migrate to SQL Server
retaining the "Front End" as the user interface. (The app' is already a "split" database
with a Front and Back end....)
I have done some initial testing on using SSMA to migrate my Access database To SQL
Server Express.
Clearly I don't understand some things and I thought I'd see if anyone has
any ideas.
Conceptually I thought that what needed to happen was that the Back End of the
database that resides on the server needed to be migrated to SQL server
and then the Front End re linked to the (now linked to SQL) tables in the Back End.
When I do this using SSMA I end up with renamed tables in the Back End
Access file that look something like "SSMA$myTableNameHere$local". I also
get the original table names underneath showing as ODBC linked tables.
So far so good.
BUT.... When I go to re-establish the linked tables from the FRONT END (The
user interface) all I can see is the "SSMA$myTableNameHere$local" names NOT
the original table names.(Now linked via ODBC)
I can link to the "SSMA,,,," tables but it would mean changing the names of
every table in every query and on every form and in all code on the Front
End! Not something I really want to do.
SO....
I thought I'd try to migrate the FRONT END and see what happens.
What I ended up with is a situation where, basically it works (there are
some serious errors and issues that I haven't even looked at yet... like
missing data etc.!!!!) and I still get the "SSMA$myTableNameHere$local"
tables and the ODBC linked tables with the original names.
I'm trying to understand...... Does this mean that we would do the
migration on the Front End and then just copy the same file to each user's
computer?
Another subject I'm a little confused about is that I can't link via ODBC
to SQL Server Express on the local machine (ie my computer) so I can't test
migrating the Back End and then linking to the tables via the Front End as I
have in the past in more of a client/server situation.
Assuming that SSMA replaces the tables in your back end with links to the SQL Server, all you need to do is delete the original table links in your front end and import the newly-created table links from the back end. You can then discard the back end, since it's not used for anything at all any longer.
I did transfer all my tables one by one to SQL Server 2005 fro Access DB back-end using ODBC.Instruction:
Open Access DB(back-end)
Right-click on table, you need to transfer
Scroll down drop-down box and select ODBC Databases
Select Data Source dialog box opened, Click "New" button
Create new data source dialog box opened
Scroll to the bottom and select SQL Server, Click Next
Give name to your Data Source, Click Next, Click Finish
Create New Data Source Dialog opens
Give some discription OR leave empty, Type Name of your SQL Server (you named it, when install SQL Server on your machine)
Click Next, Click Next
Check "change default database to check box
Select DB where you want your data transfer to
Click Next, Click Finish
NOTE: You need to create new DB (empty) on SQL Server, before doing all this
Now: Right-click any table, select Export, select from drop-down list ODBC, from Data Sources window select your Data Source, You created, Click OK
Use SQL Server with SQL Management Studio Express.
All dates must have a input mask; all text and Memo must have Allow Zero Length =Yes
After all disconnect all links from Access back-end, and establish links from SQL.RENAME all newly linked tables to old names. Use Fron-end user interfase, until do some new.
Forgive my lack of knowledge of Acronym Soup, but I assume SSMA is the SQL Server 2005 "import data wizard" or the wizard in Access to send the data to SQL Server. It appears that you sent the data to SQL Server from Access - something you don't want to do. You want to use the DTS in SQL Server (now called SSIS or something?) to import the data into SQL Server. Then you'll have your tables in SQL Server. Then, simply create your DSN entry for the SQL Server and re-link your tables. All should be well.
Overall, the general rule is to import Access tables using SQL Server instead of using Access to send the data to SQL Server.
I'd bite the bullet and rename the tables on the SQLServer side back to the friendly names that you had in the original database. You'll probably have less problems. Especially if you have any embedded code the MS Access side.
As far as how you will deploy the MS Access side now, it should be pretty much create the ODBC link on the user's workstation, and copy the MS Access file to their desktop (although you might want to make an MDE (or the 2007 equivalent) to prevent them from accidentally breaking it).
Frankly, now that you have migrated, you need to look at the design of your tables. It is my experience that the wizards for Access migration do a poor job of selecting the correct datatype. For instance if you had a memo field, you might easily get away with a varchar field instead but the last wizard I used (an earlier version) always converted them to text fields. Now would also be the time consider some fixes such as making date fileds datetime instead of character based if you have had that mistake in the past.
I would never consider using a wizard again to do data migration myself having experienced how very badly they can do it.
You will alos find that just converting the data to SQL Server is often not eough to really get any performance benefit. YOu will need to test all the queries and consider if you can convert them to stored procs instead if they are slow. Eliminating the translation from Jet SQL to T-sql can being performance improvements. Plus there are many features of t-sql that can imporve performance that do not have Access equivalents. Access is not big on performance tuning, but to get the benefit of performance tuning with a SQL Server backend, you need to have SQL Server specific queries written. INdexing needs to be considered if the Access tables were not indexed properly.
Using SSMA is different when you use odbc. If you have an application using fully access (back end and front end). You can manipulate objects easily bounding forms, using DAO, etc.. without problem, then when u need to migrate database to sql server u can use directly odbc (by linking yourself tables to sql server), ssma, ... the main problem how to preserve bounded forms, queries, code in the client-side.
If U use directly odbc you must relink by yourself all objects and change code but if u use ssma, you have to do nothing, you will continue to work as u did before. The problem with SSMA is how to deploy the front end to the clients if you developed client side in other place using another sql server?