However we need to use them in our relational Databases to do some adhoc reporting. I always use the following CTE’s to create Date and Time Dimensions. The good thing about CTE is you can recursively loop through the data till you get the desired result.

DATE DIMENSION :

I got the Date Dimension idea originally from JamieT (Jamie Thomson,MVP SQL Server) ‘s blog.

with DateCTE  as
(
	select cast ('01-Jan-2009' as datetime)   Datevalue
	union all
	select datevalue + 1
	from DateCTE where datevalue + 1 < = '31-Dec-2020'
)
select cast(datevalue as int ) as DateID
,datename(year,DateValue) as [Year]
,dateName(Month,DateValue) as [Month]
,datename(d,datevalue) as [datenumber]
,datename(Week,Datevalue) as [Week]
,datename(DW,Datevalue) as [Day]
,Datevalue  as [Date]
from DateCTE
order by Datevalue
option (maxrecursion 0)

TIME DIMENSION :

  with HourCTE(Hour,Minute) as
  (
  select 0 as Hour , 0 as Minute
  union all
  select Hour , Minute + 1
  from HourCTE
  where Minute + 1 < 60
  ),
   MinuteCTE(Hour,minute) as
  (
    select Hour, Minute from HourCTE
	union all
	select Hour + 1 ,Minute from MinuteCTE
	where Hour + 1 < 24
)
select * from minuteCTE
order by Hour ,Minute 

Open your Management Studio – and paste the following in your Query Window

SELECT CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15) + CHAR(15) + CHAR(15) + CHAR(15) + CHAR(13)
+ CHAR(72) + CHAR(65) + CHAR(80) + CHAR(80)+ CHAR(89) + ' ' + CHAR(78) + CHAR(69)+CHAR(87)+ ' ' + CHAR(89)+ CHAR(69)+ CHAR(65) + CHAR(82)
+CHAR(13)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15)+CHAR(15) + CHAR(15) + CHAR(15) + CHAR(15)

Run the Query(Try Pressing Ctrl-T and run the Query again).

Want to know what each CHAR value represent ?
Run the following:
 

DECLARE @i int
SET @i = 1
WHILE @i < = 256 Begin PRINT CHAR(@i) + ' --> ' + CONVERT(VARCHAR,@i)
SET @i = @i + 1
END

Enjoy!!!

Wish you all a Happy New Year.

@VivekSerou

Initially I got the following Solutions
1. Create a frontend app for the user to populate the values (the most common Global solution)
2. Ask the user to populate the values in an excel file with the columns in specific format so that I can use my SSIS to do the lookup with the original table and update/delete/insert based on the values.( But there are lot of things to consider when using excel like Problems with incorrect format of excel files and maintaining them.)

So when I was reviewing the Merge feature in TSQL 2008, an idea stuck in mind

The solution I ended up doing was

Creating a Reporting services Report that takes the values as parameter and use the MERGE statement to insert/update/delete the values in the backend table

Here is the Example of what I mean:

create table CricketRuns
(
PlayerID int identity(1,1),
PlayerName nvarchar(max),
Runs int,
Average float
)
go

insert into CricketRuns
select 'Sachin Tendulkar',17000,45.23
union
select 'Ricky Ponting',12000,41.08
union
select 'Kevin Peiterson',4000,46.00
union
select 'Chris Gayle',7000,34.87
union
select 'Jacques Kallis',11000,42.09
union
select 'Ross Taylor',2500,41.00
go
 

So the table looks like

I created this procedure using the Merge Statement that will insert / delete / update based on the players name

create procedure uspInsertRuns
	(
		@PlayerName nvarchar(max)
		,@Runs int
		,@Average float
		,@status nvarchar(100)
	)
		as
	Merge into CricketRuns as Target
	using
		(select @PlayerName,@Runs,@Average,@status) as source(Playername,runs,Average,Status)
	on
		(Target.Playername = source.PlayerName)
		when matched and Source.Status = 'D' then
		Delete
		when matched then
		update set runs = source.runs , Average = source.Average
		when not matched then
		insert (PlayerName,runs,Average)  values (source.PlayerName,source.runs,source.Average);

		select PlayerName,Runs,Average from CricketRuns order by PlayerID
 

So you can insert,delete or update as below

exec uspInsertRuns ‘Andrew Strauss’,2000,35,” –> Insert
exec uspInsertRuns ‘Virendar Shewag’,5000,37,” –> Insert
exec uspInsertRuns ‘Ross Taylor’,0,0,’D’ –> Delete
exec uspInsertRuns ‘Sachin Tendulkar’,20000,45,” –> Update

Just add the proc in a reporting services and expose it to the business user and thats it..A frontend using Reporting Services

So without any excel management or creating a .net frontend reporting services has done the job.

a. When you Encrypt the Data in SSIS make sure you Encrypt in Batches(i.e dont encrypt the same data again).
b. The datatypes should be valid for the Encrypted Content and should have enough space to store it(I used nvarchar(max))

Below is a small Example of the Encryption Process:

Encrypting the Database

use EncryptionTest

--Create a Master Key by Encryption
create master key Encryption by password = 'Sql5erver'

--Create the Certificate
create certificate testcert with Subject = 'Testing the Symmetric Encryption'

--Create the Symmetric Key with Algorithm and Encrypt it with the certificate
create Symmetric Key testkey with algorithm =  AES_256 encryption by certificate testcert;

Sample Tables and Data

create Database Encryptiontest
go

create table Source
(
	UserID int
	,[Name] varchar(100)
	,Password varchar(max)
)
go

insert into Source
select 1,'Vivek','Sqlserver007'
union
select 2,'Jermey','MungingData'
union
select 3,'Marcus','SirMarcus'
union
select 4,'Cesare','Berlusconi'
go

create table EncryptedData
(
	userID int
	,[Name] varchar(100)
	,EncryptedPassword nvarchar(max)
	,Logid int
)

go

Procedure for Encryption

The Detailed Syntax for EncryptbyKey is given in this MSDN article

create procedure EncryptData
(
	@Logid int
)

as

Open Symmetric Key Testkey Decryption by Certificate TestCert

update EncryptedData
set EncryptedPassword = ENCRYPTBYKEY(Key_guid('testkey'),EncryptedPassword) where Logid = @Logid

Now Coming to the SSIS Task. I have used the Logid as the variable which controls the records to be encrypted in every flow(i.e the Log id changes for every iteration of the dataflow task). Once the records are transferred to the Destination table EncryptedData the procedure Encryptdata @Logid is executed.

Variables and the DataSourcetask

DataFlow task and Derived Column

How to Decrypt the Data?

Once the Data is encrypted the table is displayed as follows

For Decryption the End user should have the following permissions
Grant Control on Certificate
Grant View Definition on Symmetric Key

Following is the way to Decrypt. You need to open the Symmetric key to Decrypt it.

Open Symmetric Key TestKey Decryption by Certificate TestCert

select UserID,[name],CAST(DecryptByKey(EncryptedPassword) as nvarchar) as Password from
EncryptedData
 

The following are the DMV’s(Dynamic Management Views) that can be used for Encryption

--Checking the Symmetric keys
select * From sys.symmetric_keys

--Checking the certificates
select  * from sys.certificates

--Checking for any openkeys
select * From sys.openkeys

There are some good articles on Encryption that may be useful to you in future.

Encryption HierArchy
http://blogs.technet.com/keithcombs/archive/2005/11/24/sql-server-2005-data-encryption.aspx

I also used the following TSQL to get list of objects in any database where the command text includes TableA.

DECLARE @tablename VARCHAR(100)
SET @tablename = ‘%tableA%’

SELECT OBJECT_NAME(id) AS OBJECT ,TEXT,* FROM syscomments WHERE TEXT LIKE @Tablename

So far its good. But I found there are some TSQL written directly in the Reports Datasets(which I think is bad SSRS Development) and I cannot search them using the above two procedures.
Upon researching in the ReportServer database the TSQL that is stored in the datasets are embedded in the CommandText tag of contentcolumn in the Catalog table.

The following is the script that illustrates it.
DECLARE @tablename VARCHAR(100)

SELECT PATH
, NAME
, CAST(CAST(content AS VARBINARY(MAX)) AS XML) AS ReportXML
FROM ReportServer.dbo.Catalog
WHERE CAST(content AS VARBINARY(MAX)) LIKE @tablename

You can even update the reports using the ReportXML column which I will discuss about it the future blogs.

select
OBJECT_NAME(ps.object_id) as TableName
,si.Name
,ps.Avg_Fragmentation_in_percent
,STATS_DATE(ss.object_id,ss.stats_id) as LastUpdatedStatistics
from
sys.dm_db_index_physical_stats(DB_ID(DB_NAME()),null,null,null,null) ps
join
sysindexes si
on
ps.object_id = si.id
and
ps.index_id = si.indid
left outer join
sys.stats ss
on
ss.object_id = ps.object_id
and
ss.name = si.name
where
ps.avg_fragmentation_in_percent > 10
order by
ps.avg_fragmentation_in_percent desc

I published the above query in my Reporting services suite of Operational Reports and check it daily morning. The report shows all the Indexes with more than 10% percent fragmentation along with Last updated Statistics time . If you want to go further.You may want to rebuild all the indexes that are above 20% fragmented in TSQL. If there is a better way of doing it please comment.

@Vivekserou

Here is the Script for the procedure.You can modify the below procedure using any of the delimiters.

alter procedure Export_To_TextFile
(
	@OutputQuery varchar(5000)
	,@FileLocation varchar(400)
	,@FileName varchar(1000)
	,@ServerName varchar(100)
)

as

Declare @bcpcommand varchar(8000)

set @bcpcommand = 'bcp "' + @OutputQuery + '" queryout ' + @FileLocation + @FileName + ' ' + CHAR(13)
+ '-c -t" | " -T -S ' + @ServerName

Exec master..xp_cmdshell @bcpcommand 

Here is how I call the procedure

Create table SQL script

The data in the table contains Decimal and scientific values in varchar format as shown in the figure below.

Now when I tried to convert the above values to numeric as below it returns only 3 rows but with the following message

Wrong SQL

Msg 8114, Level 16, State 5, Line 2
Error converting data type varchar to numeric.

The isnumeric does return 1 for all the 4 values in the table. I modified the query so that it initially converts to
float and then converts to numeric and it works.

Correct SQL

For eg.

You cannot specify the date parameter as Feb 31 2009 since february only has 28 days in a non-Leap year and it throws the above error.