2

我正在从事一个监控项目，每次收集数据时，我都会为该集合生成一个新 ID。

我想在索引视图中有最新的代码。

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

像这样的东西：

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)

这是可能的还是有其他选择？

这就是我打算使用它的方式：

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].[monit_server_space_by_drive] 的主键由三个字段组成：

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]

以下是此表包含的数据的示例。每次我为连接收集数据时，THE_ID 都会增加一。在同一个 THE_ID 中，我拥有环境中的所有服务器。

在此处输入图像描述

4

2 回答 2

1

THE_ID 是否定义为 [dbo].[monit_server_space_by_drive] 表上的唯一聚集索引？字段是身份吗？如果是这样，我认为您不会通过使用索引视图来实现任何收益。

比较这些可能性的性能：

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

或者

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;

或者

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。

于 2014-05-12T15:40:00.457 回答
1

如果这可行，那么每次插入新记录时，都必须更新视图的索引。

为什么不更直接地使用触发器将最后一个 ID 更新到一个小表中呢？

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，您可以按预期使用此表而不是视图。

于 2015-04-22T13:58:05.620 回答