使用索引视图获取最新记录

时间:2014-05-12 13:48:41

标签: sql-server sql-server-2008-r2 views indexed-view

我正在开发一个监控项目,每次收集数据时,我都会为该集合生成一个新的ID。

我希望在索引视图中获得最新代码。

SELECT THE_ID = MAX(THE_ID)  FROM [dbo].[monit_server_space_by_drive]

类似的东西:

CREATE VIEW V_LAST_ID 
WITH SCHEMABINDING
AS
SELECT THE_ID = MAX(THE_ID)  FROM [dbo].[monit_server_space_by_drive]

CREATE UNIQUE CLUSTERED INDEX UC_IDX_V_LAST_ID  ON V_LAST_ID (THE_ID)

这可能还是有其他选择?

这就是我打算如何使用它:

SELECT 
   M.[THE_ID]
  ,M.[serverid]
  ,s.SQLServerName
  ,s.[Environment]
  ,M.[DRIVE]
  ,M.[Volume_Size_GB]
  ,[VolumeUsed_GB] = CAST (M.[VolumeUsed_MB] / 1024.00 AS NUMERIC (18,2))
  ,M.[VolumeSpaceAvailable_GB]
 , DrivePercentUsed = 100 - M.[VolumePercentAvailable]
  ,M.[VolumePercentAvailable]

  FROM [dbo].[monit_server_space_by_drive] M
  INNER JOIN DBO.V_LAST_ID  V 
      ON M.THE_ID = V.THE_ID
  INNER JOIN [dbo].[tblServers] S
      ON M.serverid = s.ServerID
  order by M.[ClusterName]

表[dbo]的主键。[monit_server_space_by_drive]由三个字段组成:

USE [Monitoring]
GO

ALTER TABLE [dbo].[monit_server_space_by_drive] ADD  CONSTRAINT [PK_monit_server_space_by_drive] PRIMARY KEY CLUSTERED 
(   [THE_ID] ASC,
[serverid] ASC,
[volumeID] ASC
)ON [FGMONITORING]

表格定义:

CREATE TABLE [dbo].[monit_server_space_by_drive](
[THE_ID] [int] NOT NULL,
[serverid] [int] NOT NULL,
[ClusterName] [nvarchar](260) NOT NULL,
[nodeid] [int] NOT NULL,
[DRIVE] [nchar](1) NULL,
[volumeID] [int] NOT NULL,
[LastSync] [datetime] NULL,
[Volume_Size_MB] [numeric](18, 2) NULL,
[Volume_Size_GB] [numeric](18, 2) NULL,
[VolumeSpaceAvailable_MB] [numeric](18, 2) NULL,
[VolumeSpaceAvailable_GB] [numeric](18, 2) NULL,
[VolumePercentAvailable] [numeric](5, 2) NULL,
[dt] [datetime] NULL,
[_year] [smallint] NULL,
[_month] [tinyint] NULL,
[_day] [tinyint] NULL,
[_week] [tinyint] NULL CONSTRAINT [DF_monit_server_space_by_drive__week]  DEFAULT ((0)),
[_hour] [tinyint] NULL,
[_min] [tinyint] NULL,
[_weekday] [tinyint] NULL,
[VolumeUsed_MB]  AS ([volume_size_MB]-[volumeSpaceAvailable_MB]) PERSISTED,
 CONSTRAINT [PK_monit_server_space_by_drive] PRIMARY KEY CLUSTERED 
 (
[THE_ID] ASC,
[serverid] ASC,
[volumeID] ASC
) ON [FGMONITORING_INDEX]
) ON [FGMONITORING]

以下是此表所包含数据的示例。每次我收集连接数据时,THE_ID都会增加一个。在同一个THE_ID中,我拥有环境中的所有服务器。

enter image description here

2 个答案:

答案 0 :(得分:1)

THE_ID是否定义为[dbo]。[monit_server_space_by_drive]表中的唯一聚簇索引?这个领域是一个身份吗?如果是这样,我认为你不会通过使用索引视图来实现任何收益。

比较这些可能性的表现:

Set Statistics IO ON;
Set Statistics Time ON;
Select * from V_LAST_ID;
Set Statistics IO OFF;
Set Statistics Time OFF;

OR

Set Statistics IO ON;
Set Statistics Time ON;
SELECT THE_ID = MAX(THE_ID)  FROM [dbo].[monit_server_space_by_drive];
Set Statistics IO OFF;
Set Statistics Time OFF;

OR

Set Statistics IO ON;
Set Statistics Time ON;
SELECT Top 1 THE_ID FROM [dbo].[monit_server_space_by_drive] Order by THE_ID Desc
Set Statistics IO OFF;
Set Statistics Time OFF;

在ssms中单独运行每个,消息选项卡将显示cpu和持续时间。

我怀疑这两种替代解决方案的性能至少与索引视图一样好,但您需要自己测试它们。

答案 1 :(得分:1)

如果这样做,那么每次插入新记录时,都必须更新视图的索引。

为什么不直接使用触发器将最后一个ID更新到一个小表中。

CREATE TABLE DBO.V_LAST_ID( [THE_ID] [int] NOT NULL );

INSERT DBO.V_LAST_ID (THE_ID) VALUES (0); --Starting row

CREATE TRIGGER Dbo.KeepV_LAST_IDNumber ON [dbo].[monit_server_space_by_drive]
AFTER INSERT AS
BEGIN
   DECLARE @MaxInserted int;

   SELECT @MaxInserted = MAX(THE_ID) FROM inserted;

   UPDATE DBO.V_LAST_ID
   SET THE_ID = @MaxInserted
   WHERE THE_ID < @MaxInserted;
END

然后,您可以按照预期使用此表而不是视图。