Tag: Sequences

SQL Server “Denali” – SEQUENCEs–Continued…


After reading my previous article about SEQUENCE one of my friend asked a question about the performance of using Sequence for quick and continuous generation and whether there is a way to create and manage this using Management Studio ?

Yes, SEQUENCE support CACHE option

  1. /* Create Sequence with Cache */
  2.  
  3. CREATE SEQUENCE RunningNumbers
  4. As INT
  5. MINVALUE 1
  6. NO MAXVALUE
  7. START WITH 1
  8. CACHE 500
  9. NO CYCLE;

 

Using CACHE option, you can specify how many values should be pre generated and available ready, you can set the cache size based on your application’ requirement.

Default cache size is 20

To use in Loop for quick processing

To get a range of numbers to use in a loop, you can use sp_sequence_get_range System stored procedure, so that you can get a range and not have to use NEXT VALUE option for each loop call

  1. /* Get Range of Numbers using Sequence */
  2. DECLARE
  3. @RANGE_FIRST_VALUE SQL_VARIANT,
  4. @RANGE_LAST_VALUE SQL_VARIANT,
  5. @RANGE_CYCLE_COUNT INT,
  6. @SEQUENCE_INCREMENT SQL_VARIANT,
  7. @SEQUENCE_MIN_VALUE SQL_VARIANT,
  8. @SEQUENCE_MAX_VALUE SQL_VARIANT;

 

  1. EXEC sp_sequence_get_range
  2. @sequence_name = 'RunningNumbers',
  3. @range_size = 15,
  4. @range_first_value = @range_first_value OUTPUT,
  5. @range_last_value = @range_last_value OUTPUT,
  6. @range_cycle_count = @range_cycle_count OUTPUT,
  7. @sequence_increment = @sequence_increment OUTPUT,
  8. @sequence_min_value = @sequence_min_value OUTPUT,
  9. @sequence_max_value = @sequence_max_value OUTPUT;

  1. SELECT
  2. 'RunningNumbers' AS [Sequence Name],
  3. @range_first_value AS [Sequence First Value],
  4. @range_last_value AS [Sequence Last Value],
  5. @range_cycle_count AS [Range Cycle Count],
  6. @sequence_increment AS [Sequence Increment],
  7. @sequence_min_value AS [Sequence Min Value],
  8. @sequence_max_value AS [Sequence Max Value];

image

You can use [Sequence First Value] and [Sequence Last Value] and use it in your loop to insert data.

How to create and manage SEQUENCE from Management Studio ?

Step 1 : Open SQL Server “Denali” Management Studio, and connect to SQL Server “Denali” instance and go to your database in Object Explorer

Step 2: Expand “Programmability” and then expand "Sequences"  to currently available ones

image

Step 3: To create a new sequence, right click on Sequence and click on “New Sequence”

image

Step 4: Specify the name and other required options such as Start Value, Increment By, etc.. and click “Ok” to create Sequence

image

Step 5: To modify the Sequence, right click on the selected sequence and click “Properties”

image

If you see the above screenshot, you might notice that Data Type, Precision or name of sequence can not be modified. You can change the other values, similarly you will see an option to restart the sequence as well.

Note: To rename you can use the “Rename” option in Context menu.

Step 6: To drop the sequence, right click on the selected sequence and click “Delete”

image

Let me know your experiences and feedback about SEQUENCEs

Advertisement

SEQUENCE–How to create Identity Key across tables in SQL Server “Denali” ?


SQL Server “Denali” introduces a new feature called “SEQUENCE”. SEQUENCE is a user-defined object that generates a sequence of numeric values according to specified options.

Its not bound to a table like IDENTITY, You can use SEQUENCE across tables

Supported Data Types for creating SEQUENCE are as follows:

  • TinyInt
  • SmallInt
  • Int
  • BigInt
  • Decimal
  • Numeric
  1. /* Create Sequence With Start Values */
  2. CREATE SEQUENCE RunningNumbers
  3.     START WITH 1
  4.     INCREMENT BY 1;
  5. GO

 

The above code creates a simple SEQUENCE with starting number as 1 and increments by 1

To get the next number in SEQUENCE, you need to use “NEXT VALUE” option

  1. SELECT (NEXT VALUE FOR RunningNumbers) AS [RunningNumbers];

image

The below code might syntactically look same as the SEQUENCE "- RunningNumbers, but if you look at the output, its entirely different

  1. /* Create Sequence Without Start Values */
  2. CREATE SEQUENCE RunningNumbers2
  3.     INCREMENT BY 1;
  4. GO

  1. SELECT (NEXT VALUE FOR RunningNumbers2) AS [RunningNumbers2];

If you see the output from RunningNumbers2, you will be getting a Negative value, its due to not specifying the Start number, so make sure you specify a Positive Start Number if you don’t want your table keys to Negative

image

Other options available in Creating Sequence are as follows:

  1.  
  2. CREATE SEQUENCE RunningNumbers3
  3.    AS tinyint
  4.     START WITH 1
  5.     INCREMENT BY 1
  6.     MINVALUE 1
  7.     MAXVALUE 5
  8.     CYCLE ;
  9. GO

 

You can specify Minimum Value and Maximum Value and whether to Cycle the numbers when reaching maximum value or not

 

  1. SELECT NEXT VALUE FOR RunningNumbers3 AS ID, Name FROM sys.objects ;
  2. GO

 

If you see the below output, for every 5 records, the IDs are recycling and starts from 1 again

image

Restarting the Sequence

To restart the Sequence, you need to ALTER the sequence with RESTART option and start value

  1. /* Restart the Sequence */
  2.  
  3. ALTER SEQUENCE [dbo].[RunningNumbers2]
  4. RESTART WITH 1

  1.  
  2. SELECT (NEXT VALUE FOR RunningNumbers2) AS [RunningNumbers2];

Since we restarted the Sequence with 1, Now we are getting output as 1 for RunningNumber2

image

Find the available Sequences in a database

  1. SELECT
  2.     object_id, name, type_desc, start_value, current_value, increment,
  3.     minimum_value, maximum_value, is_cycling, is_cached, cache_size
  4. FROM sys.sequences

You can Sys.Sequence system view to query the list of available Sequences in a database

image

How to use it as Identity across tables ?

Create tables for each reqion

  1. — Create tables
  2. CREATE TABLE Orders_West
  3.     (OrderID int PRIMARY KEY,
  4.     Name varchar(20) NOT NULL,
  5.     Qty int NOT NULL);
  6. GO
  7.  
  8.  
  9. CREATE TABLE Orders_East
  10.     (OrderID int PRIMARY KEY,
  11.     Name varchar(20) NOT NULL,
  12.     Qty int NOT NULL);
  13. GO
  14.  
  15. CREATE TABLE Orders_South
  16.     (OrderID int PRIMARY KEY,
  17.     Name varchar(20) NOT NULL,
  18.     Qty int NOT NULL);
  19. GO
  20.  
  21. CREATE TABLE Orders_North
  22.     (OrderID int PRIMARY KEY,
  23.     Name varchar(20) NOT NULL,
  24.     Qty int NOT NULL);
  25. GO

 

Insert data to table using Sequences

  1. — Insert Four records
  2. INSERT Orders_West (OrderID, Name, Qty)
  3.     VALUES (NEXT VALUE FOR RunningNumbers2, 'Apples', 2) ;
  4. INSERT Orders_East (OrderID, Name, Qty)
  5.     VALUES (NEXT VALUE FOR RunningNumbers2, 'Oranges', 1) ;
  6. INSERT Orders_South (OrderID, Name, Qty)
  7.     VALUES (NEXT VALUE FOR RunningNumbers2, 'Grapes', 1) ;
  8. INSERT Orders_North (OrderID, Name, Qty)
  9.     VALUES (NEXT VALUE FOR RunningNumbers2, 'Banana', 1) ;
  10.  
  11. GO

 

Query the inserted data

  1. SELECT * FROM Orders_West
  2. UNION
  3. SELECT * FROM Orders_East
  4. UNION
  5. SELECT * FROM Orders_North
  6. UNION
  7. SELECT * FROM Orders_South

 

If you see the Union output from 4 different tables, we are able to create a identity column across 4 tables, This will be very helpful when you horizontally partition data across tables

image

Delete the Sequences

  1. DROP SEQUENCE RunningNumbers;
  2. DROP SEQUENCE RunningNumbers2;

I hope you all find this information about Sequences useful and informative !!!

%d bloggers like this: