Comments on: SQL SERVER – 2005 – Database Table Partitioning Tutorial – How to Horizontal Partition Database Table

By: VIRAL JAIN

VIRAL JAIN — Fri, 17 May 2013 10:45:37 +0000

Yeah… .NDF – Secondary database file used by Microsoft SQL Server; created when the primary .MDF database becomes too large (such as when there is no disk space left); may also be manually added to the SQL Server File Group to distribute data between the two files.

By: laxman

laxman — Sat, 06 Apr 2013 12:08:54 +0000

Hi Pinal

Please can you reply on my above question on multiple column partition in a table.

By: laxman

laxman — Tue, 02 Apr 2013 18:05:38 +0000

Hi Pinal

Is it possible to partition a table using 2 column value’s instead of only 1 column for the partition function?

Consider a table with 4 columns

ID (int, primary key,
name (varchar),
Country (varchar),
region (varchar)

ex:
ID name Country region

1 rahul India karnataka
2 raj India maharastra
3 brad Usa california
4 richard Usa arizona

I want partition data combining based on country and region
1st partition data like india and karnataka
2nd partition data like india and maharastra.
3rd partition data like usa and california
4th partition data like usa and arizona

By: SQL SERVER – Weekly Series – Memory Lane – #013 « SQL Server Journey with SQL Authority

Sat, 26 Jan 2013 01:31:16 +0000

[...] Database Table Partitioning Tutorial – How to Horizontal Partition Database Table Database partitioning is a very important concept and very useful. Many a times DBA partition database for performance and many times to handle large volume. As this operation is usually done when there are huge amount of data exists this is quite a critical task to be performed by the DBA. In this blog post, I have explained step by step how to create horizontal partition which provides optimal performance. [...]

By: manoj

manoj — Mon, 21 Jan 2013 13:53:33 +0000

Manoj

I am new to DB, I have a database with 30gb single mdf file, and the perfromance of the database appears to be slow.

I have below queries,

1.how can i split the 30gb mdf file into many mdf files?, if this can be done, will this splitting spread the 30gb data accross multiple mdf/ndf files?

2.I have tried adding the mdf/ndf files ,under database->properties->files, if Iuncheck the autogrowth of the 30gb mdf file, will the data automatically move to the newly added mdf/ndf files?

3.can something be done in SQL, that when importing the .bak file of 30gb, data can be split accross multiple mdf/ndf files created prior to the restoration of the database backup file?

Please help me.. Any help ASAP, will be a lifesaver for me

By: this website

this website — Tue, 15 Jan 2013 22:48:26 +0000

Hey There. I found your blog using msn. This is a really well written article.
I’ll be sure to bookmark it and come back to read more of your useful information. Thanks for the post. I’ll definitely return.

By: akondo

akondo — Thu, 10 Jan 2013 14:51:46 +0000

It is .ndf

By: Vivek

Vivek — Thu, 10 Jan 2013 11:50:15 +0000

Hi Dave, can you explain me how can I partition my data warehouse table it has got one clustered composite primary key [snapshot_date_key, process_date_Key, process_unit_key, metric_key, uom_key, msr_type_key] it has got around 14 million data and all its queries are so effective. we’d like to tune it.. hence we are looking for partitioning.
kindly share your insights and thoughts about this? also revert me in case if you need additional information.
-Vivek

By: cornasdfcornasdf

cornasdfcornasdf — Mon, 17 Sep 2012 09:53:45 +0000

Lian, We just went through this recently (see blog below for more details).
When you apply your partition function sql server will reorganize the data into files to match your function. be aware that this can take some time!

As new data is written into the database, is is written to the correct file/partition.

By: Lian

Lian — Mon, 17 Sep 2012 09:16:26 +0000

Pinal, this process is fine for new tables , but how do I create a partition on an existing table and how is the new partion updated regularly ?

By: Chris

Chris — Wed, 29 Aug 2012 14:53:06 +0000

Hi Dave, thanks for the simple tutorial. I just have a question which I’m not sure if you have answered it already. How does Partition help in performance when you have a normalized schema with 1 main table and 4 child tables all connected with a unique id (bigint).

basically on our schema the important thing is a datetime field named WEEK because this determines the chronological order of records however this field is only present on the Main/Parent table. How does partitioning affect all our Query statements where they are joined together with the child tables?

Thank you in advance.

By: Jimmy

Jimmy — Fri, 24 Aug 2012 16:40:42 +0000

Very nice simple example. Thanks Dave.

By: sap

sap — Thu, 23 Aug 2012 13:58:07 +0000

if i 2 partition(p1,p2) i want to truncate only p2 partition as it will populated with older data so ,what will be the sql condition to truncate only parition p2 on that table.

By: madhivanan

madhivanan — Thu, 23 Aug 2012 06:20:15 +0000

Create a procedure that deletes old data based on date column and schedule that procedure to run periodically

By: sap

sap — Wed, 22 Aug 2012 19:46:21 +0000

i have very big table which i want to implement partition as you explained but how should i truncate the data which is older than some period as automatic job

By: Andrea

Andrea — Thu, 14 Jun 2012 22:06:11 +0000

Here is an interesting question for you Dave,
I have a partitioned table with a clustered index on start_dt which is a datetime type field.
The same table has a few more datetime fields one of them being customer_call_start_date.
When creating an index on this table and having a key on the customer_call_start_date is it better to write it this way:
CREATE NONCLUSTERED INDEX [myIndex] ON [Schema].[table]
(
[Column1] ASC,
[customer_call_start_date] ASC
)
ON [Partition_Scheme_Partition]([START_DT])

or this way:
CREATE NONCLUSTERED INDEX [myIndex] ON [Schema].[table]
(
[Column1] ASC,
[customer_call_start_date] ASC
)
ON [Partition_Scheme_Partition]([customer_call_start_date])

Thank you
Andrea

By: Hirdesh Sharad

Hirdesh Sharad — Sun, 03 Jun 2012 06:45:47 +0000

Dear pinal,
I am regularly reading your sql blogs,
the main problem in your blog is you are not specifying that which edition of sql server contain the particular concept. ex: partitioning is possible only in enterprise edition but you didnt mention it anywhere,
So pls form next time keep in mind to tell edition of sql server.

By: Tharindu Dhaneenja

Tharindu Dhaneenja — Mon, 28 May 2012 05:16:15 +0000

Hi Pinal, What are Disadvantages having table Partition? is Insert,Update and delete slow?
Thanks
Tharindu Dhaneenja

By: Ramesh

Ramesh — Wed, 23 May 2012 10:43:28 +0000

Hi Dave,

I am created partitioned table using datetime,
I defined range for month upto one year ’2011-01-01′,’2011-02-1′,’2011-03-1′
And I created a Partitioned Table.

when I Looking into sys.partitions

SELECT partition_id, object_id, partition_number, rows
FROM sys.partitions
WHERE object_id = OBJECT_ID(‘partitionedtablet’)
order by partition_Number

______________________________________________________________

*** I am getting twise or thrise of the partion_number of same number,
How and why??

And some having rows and other Partition row of same number does not have rows, but index column saying 0,1 and 2.

Can anyone explain please……………..

Ramesh Chandler.

By: cornasdf

cornasdf — Mon, 14 May 2012 11:51:11 +0000

Thanks for this excellent article (and several others). I have used this as a base for a walthrough on backup and restores of partitioned databases. http://cornasdf.blogspot.com/2012/05/basic-walkthrough-of-partial-backup-and.html

By: Alia

Alia — Mon, 30 Apr 2012 20:48:33 +0000

Hi Dave,
How Can I insert records on daily basis in a partitioned table.
A data warehouse table which I update on daily basis now I created five partition on it one for each year and the last partition is empty.
I need to know the best practice in this regard.
Alia

By: Vivekananda Swamy

Vivekananda Swamy — Wed, 11 Apr 2012 15:49:21 +0000

Hi Dev,
We have 3 years of data in more than 30 tables based on Date partition. I mean we are creating one partion for each business day. Now we want the purge the data which is 18 months older. How do I efficiently purge the data ? Can you please help us on it in SQL Server 2005.

By: Sampong Kwame

Sampong Kwame — Fri, 02 Mar 2012 21:11:03 +0000

I am new to using SQL server but I know with your help I will understand it fast. Please can you send me something because am a beginner and I really want to learn. Thank you.

By: SQL SERVER – Advantages of Partitioning – Quiz – Puzzle – 30 of 31 « SQL Server Journey with SQL Authority

Tue, 31 Jan 2012 01:31:56 +0000

[...] Server Interview Questions and Answers ISBN: 1466405643 Page#168-170 Database Table Partitioning Tutorial – How to Horizontal Partition Database Table White Paper – Partitioned Table and Index Strategies Using SQL Server [...]

By: Sivakumar

Sivakumar — Tue, 27 Dec 2011 15:06:02 +0000

It is very helpful and useful guide for beginners. Excellent Job.

Thanks Pinal. Keep it up.