InstantForum.NET v2.9.0SQLServerCentralhttp://www.sqlservercentral.com/Forums/notifications@sqlservercentral.comTue, 18 Jun 2013 22:44:36 GMT20http://www.sqlservercentral.com/Forums/Topic1394392-391-1.aspx[quote][b]mah_j (12/10/2012)[/b][hr]also i think the page split would be decreased in the partition that i want to select from it.[/quote] NO , partitoning is not related with page spliting and for partitiioning see this [url]http://www.brentozar.com/archive/2012/03/how-decide-if-should-use-table-partitioning/[/url]Mon, 10 Dec 2012 03:06:03 GMTBhuvneshhttp://www.sqlservercentral.com/Forums/Topic1394392-391-1.aspxIf you are partitioning with the idea that queries are automatically faster on partitioned tables, don't waste your time.Mon, 10 Dec 2012 02:44:18 GMTGilaMonsterhttp://www.sqlservercentral.com/Forums/Topic1394392-391-1.aspxThere is a sharply increase in the size of transactions that is inserted in that table,and if i can do partitioning the remaining range of kind field will be half,and the performance of select statements on the table will be increased,also i think the page split would be decreased in the partition that i want to select from it.Mon, 10 Dec 2012 02:36:35 GMTmah_jhttp://www.sqlservercentral.com/Forums/Topic1394392-391-1.aspxFirst question. Why are you partitioning? What's the purpose, what's the goal?Sun, 09 Dec 2012 08:50:56 GMTGilaMonsterhttp://www.sqlservercentral.com/Forums/Topic1394392-391-1.aspxHi I have a large table which there is just select and insert on it.It has a field named 'TranType'(int) and for some of the types we just have insert and no select at all.I want to use partitioning to separate this data into another file.but there is not any particular range for values . For example the types with no select are:1004,1006,1007,1009,1010,2000How can i use the CREATE PARTITION FUNCTION in this case?CREATE PARTITION FUNCTION PartitionDB_PartitionRange (INT)AS RANGE LEFT FORVALUES (????);Another basic question about partitioning:I have a table like thisCREATE TABLE [dbo].[Trans]( [Num] [int] NOT NULL, [Date] [datetime] NOT NULL, [ID] [bigint] NULL, [Kind] [int] NULL, [Success] [bit] NULL, [CNum] [nvarchar](20) NULL, [Data] [nvarchar](4000) NULL, [Counter] [int] NOT NULL, [IP] [nvarchar](25) NULL, [Result] [int] NULL, CONSTRAINT [PK_Trans] PRIMARY KEY CLUSTERED([Num] ASC, [Date] ASC) WITH ( PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON ) ON [PRIMARY]) ON [PRIMARY]so it has a Clustered Index For PRIMARY KEY and 6 NonClustered Indexes.for partitioning on the feild "kind" should i drop all the FK and PRIMARY KEY,and create a new CLUSTERED Index on kind feild?there isnt any way that i dont have to drop all this indexes and FK s?Sun, 09 Dec 2012 03:54:29 GMTmah_j