SQL*_*pro 9 sql-server dump tempdb sql-server-2019
我们在 2 台新物理机上的新 SQL Server 2019 有一个非常奇怪的问题：
基础结构：在 2 台新物理机（用于 AlwaysOn 副本）上开始新安装 SQL Server 2019 Enterprise（Windows Server 2019 Standard 10.0 / Build 17763 上的 15.0.2000.5 / X64）。新机器是联想：
- ThinkSystem SR630 – [7X02CTO1WW]
- 1 个 CPU：1 个 Xeon Gold 6208U – 2.90 Ghz（16 核 x 2 – 超线程）
- 256 Gb 内存 (32 Gb x 8)
该问题系统地出现在 2 台新机器上......
测试：产生故障的测试如下：
正是最后一个查询（执行了近 10 次）经常导致出现以下消息的错误：
消息 601，级别 12，状态 1，行……由于数据移动，无法使用 NOLOCK 继续扫描。
当然，我们从未实现过NOLOCK提示或READ UNCOMMITTED隔离级别。但是……信息记录在 SQL Server 事件日志中：
“缓冲区锁存器”消息示例：
等待缓冲区锁存器时发生超时 — 类型 4，bp 00000292CE3D60C0，第 9 页：18634，统计信息 0x10b，数据库 ID：2，分配单元 ID：422212527063040/140737550352030400x0x00x700x00x0702000000000000000000000000000000000007拥有任务 0x00000292AB06B848。没有继续等待。等待缓冲区锁存器时发生超时 — 类型 4，bp 00000292CE398340，第 6 页：10372，统计信息 0x10b，数据库 ID：2，分配单元 ID：422212527063040/1407375503520308400x70x00x7020000000000x10000000000x1000拥有任务 0x00000292AB073468。没有继续等待。等待缓冲区锁存器时发生超时 — 类型 4，bp 00000292CE3DC480，第 9 页：18655，统计信息 0x10b，数据库 ID：2，分配单元 ID：422212527063040/140737550352000200x3082020 秒标志，30 秒 30 秒，任务 拥有任务 0x00000292AB07B468。没有继续等待。
我们的调查：
一些似乎可以减少这种现象发生的元素：
一些似乎可以解决问题的元素：
我们移除 RAM 并将其替换为另一个，并使用不同的参考。问题依旧...
也许这与 SQL Server 2019 的新“并发 PFS 更新”有关……是否有任何跟踪标志禁用了这种新行为？
========== SCRIPT ==========
-- 1) creating database
CREATE DATABASE DB_BENCH
GO
DECLARE @SQL NVARCHAR(max) = N'';
SELECT @SQL = @SQL + N'ALTER DATABASE DB_BENCH MODIFY FILE (NAME = ''' + name + N''', SIZE = 10 GB, FILEGROWTH = 64 MB);'
FROM DB_BENCH.sys.database_files;
SET @SQL = @SQL + N'ALTER DATABASE DB_BENCH SET RECOVERY SIMPLE;'
EXEC (@SQL);
GO
USE DB_BENCH
GO
-- 2) creating table and view
CREATE TABLE T_TIME_INTERVAL_TIV
(TIV_ID INT NOT NULL IDENTITY PRIMARY KEY,
TIV_GROUP INT,
TIV_DEBUT DATETIME2(0),
TIV_FIN DATETIME2(0))
GO
CREATE VIEW V
AS
SELECT TIV_GROUP AS id, TIV_DEBUT AS intime, TIV_FIN AS outtime
FROM T_TIME_INTERVAL_TIV
GO
-- 3) inserting datas
TRUNCATE TABLE T_TIME_INTERVAL_TIV;
GO
BULK INSERT T_TIME_INTERVAL_TIV
FROM "C:\DATA_SQL\intervals.txt"
WITH (KEEPIDENTITY , FIELDTERMINATOR = '\t',
ROWTERMINATOR = '\n');
GO
-- 4) testing
SET NOCOUNT ON;
GO
SET STATISTICS TIME ON;
GO
WITH T1 As
(SELECT id, intime
FROM V
UNION ALL
SELECT id, outtime FROM V),
T2 As
(SELECT ROW_NUMBER() OVER(PARTITION BY id ORDER BY intime) NN, id, intime
FROM T1 T1_1),
T3 As
(SELECT T2_1.NN - ROW_NUMBER() OVER(PARTITION BY T2_1.id ORDER BY T2_1.intime,T2_2.intime) NN1,
T2_1.id, T2_1.intime intime, T2_2.intime outtime
FROM T2 T2_1
INNER JOIN T2 T2_2
ON T2_1.id=T2_2.id
And T2_1.NN=T2_2.NN-1
WHERE EXISTS (SELECT *
FROM V S
WHERE S.id=T2_1.id
AND (S.intime < T2_2.intime AND S.outtime>T2_1.intime))
OR T2_1.intime = T2_2.intime)
SELECT id, MIN(intime) intime, MAX(outtime) outtime
FROM T3
GROUP BY id, NN1
ORDER BY id, intime, outtime;
========== END OF SCRIPT ==========
执行脚本的文件：
2020-09-14UTC16:00，补充测试：
在 SQL Server 2019 上使用 2017 版本的数据库，如下所示：
USE master
GO
ALTER DATABASE DB_BENCH SET COMPATIBILITY_LEVEL = 140;
GO
执行相同查询超过 180 次，结果没有任何错误
使用 2019 版本并启用 TF 3925 或 3972或 3925 和 3972 的组合，导致所有执行（使用 MAXDOP 0、超线程和 NUMA 软）时出现系统错误。
小智 0
我在这里提供详细且完整的建议答案。发帖者需要按已接受的方式做出回应或提供反馈以改进解决方案。我的观点是，报告的问题不是 Microsoft SQL Server 2019 的故障，不是 TRACE FLAGS 的故障，也不是 TEMPDB 的故障。问题在于代码和需求的概念化。要求相当简单，如下所示。
这里的一般概念类似于 FILO 堆栈（先进后出）。它精确地模拟一千 (1000) 个数字分类的“演员”带着可写日记或记录卡走上拥挤的舞台，演员们在一系列百老汇演出中多次进入舞台并离开舞台，由早在新冠疫情爆发之前，丽莎·米内利 (Liza Minelli) 就在美国纽约。
1000 名演员中每一位的退出和进入循环可能会持续数天甚至数年，因为这是一场非常成功的演出，他们的退出和进入会被跟踪到当天的第二天。数据的粒度为每天 86400 秒，没有分数。我没有分析是否有多个Actor可以同时退出和进入，但数据中会有证据。
每个演员从舞台右侧进入，然后从舞台左侧退出。每个演员都用一张编号卡“标记”。Liza Minelli 的一些业务分析师将其称为“分类器”。分析师将这个徽章编号分配给每个演员，并将臭气熏天的徽章贴在他们的肩膀上。
演员的徽章编号并不唯一标识海报源数据 (SQL Pro) 中的每一行，因为具有徽章编号的演员多年来可能多次进入舞台并退出舞台，就像符号先知一样。
因此，它们的分类器在每一行上出现多次，具有不同的进入和退出日期时间。他们在舞台右侧输入的每个日期时间，演员都会在他们携带的“日期时间输入”栏中的记录卡上标记一个新的日期时间。接下来，当他们随后在舞台左侧退出时，演员会在名为“Datetime OUT”的另一列中获得另一个标记。
多年来，演员可以多次进出舞台，但 Liza Minelli 领导下的安全部门的年度报告只需要看到第一次“进入舞台右侧”，即“Datetime IN”和最后一次“退出舞台” Left”，称为“Datetime OUT”。他们需要为每一个肩上带有编号的臭徽章的演员提供这个。请注意，徽章确实很臭。事实上，即使演员们说他们不需要臭徽章，他们仍然很臭。现在咯咯笑以保持快乐。
而且，第一个将在最后，最后一个将在最前，因为丽莎·米内利有信心（见马太福音 19:30）。无论如何，以下是完成此要求/紧急情况的步骤：
这里列出了所有的代码，包括所有的DDL（数据定义）和DML（数据操作/T-SQL），甚至还有几十行结果。下面显示了方括号，而不是创建的对象名称中的字面意思。我只是将它们用作惯例。下面的代码块将由以下编号列表引用。
报告应具有以下特点：
这实际上是一个经典的STACK解决方案。先进后出（推入/弹出）。海报似乎每个分类器需要 1 行，其中该行显示分类器编号（非键整数）及其入口和出口。有关演员进入和退出人生舞台的更多信息，请参见莎士比亚。
/*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- Purpose: Creates the ActorStage Database for SQL Pro
--------------------------------------------------------
-- SQL Server 2019 Version
-- The folders must exist as shown
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
USE [master]
go
EXEC sp_configure 'show advanced options', 1;
GO
EXEC sp_configure'xp_cmdshell', 1;
GO
RECONFIGURE
-- Next, directories, will only work once. Look at your system.
GO
xp_cmdshell 'MKDIR C:\OLTP_DATA\ActorStage\System\';
GO
xp_cmdshell 'MKDIR C:\OLTP_DATA\ActorStage\Hub\';
GO
xp_cmdshell 'MKDIR C:\OLTP_DATA\ActorStage\Metadata\';
GO
xp_cmdshell 'MKDIR C:\OLTP_DATA\ActorStage\Indices\';
GO
xp_cmdshell 'MKDIR D:\OLTP_TLOG\ActorStage\';
GO
USE [master]
GO
CREATE DATABASE [ActorStage]
CONTAINMENT = NONE
ON PRIMARY -- The Primary Filegroup should never have user data on it.
( NAME = N'ActorStage_SYSDATA_0', FILENAME = N'C:\OLTP_DATA\ActorStage\System\ActorStage_SYSDATA_0.MDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
( NAME = N'ActorStage_SYSDATA_1', FILENAME = N'C:\OLTP_DATA\ActorStage\System\ActorStage_SYSDATA_1.MDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
FILEGROUP [HUB]
( NAME = N'ActorStage_HUB_0', FILENAME = N'C:\OLTP_DATA\ActorStage\Hub\ActorStage_HUB_0.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 12MB),
( NAME = N'ActorStage_HUB_1', FILENAME = N'C:\OLTP_DATA\ActorStage\Hub\ActorStage_HUB_1.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 12MB),
( NAME = N'ActorStage_HUB_2', FILENAME = N'C:\OLTP_DATA\ActorStage\Hub\ActorStage_HUB_2.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 12MB),
( NAME = N'ActorStage_HUB_3', FILENAME = N'C:\OLTP_DATA\ActorStage\Hub\ActorStage_HUB_3.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 12MB),
FILEGROUP [METADATA]
( NAME = N'ActorStage_METADATA_0', FILENAME = N'C:\OLTP_DATA\ActorStage\Metadata\ActorStage_METADATA_0.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
( NAME = N'ActorStage_METADATA_1', FILENAME = N'C:\OLTP_DATA\ActorStage\Metadata\ActorStage_METADATA_1.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
FILEGROUP [INDICES]
( NAME = N'ActorStage_INDICES_0', FILENAME = N'C:\OLTP_DATA\ActorStage\Indices\ActorStage_INDICES_0.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
( NAME = N'ActorStage_INDICES_1', FILENAME = N'C:\OLTP_DATA\ActorStage\Indices\ActorStage_INDICES_1.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
( NAME = N'ActorStage_INDICES_2', FILENAME = N'C:\OLTP_DATA\ActorStage\Indices\ActorStage_INDICES_2.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB),
( NAME = N'ActorStage_INDICES_3', FILENAME = N'C:\OLTP_DATA\ActorStage\Indices\ActorStage_INDICES_3.NDF' , SIZE = 25MB , MAXSIZE = UNLIMITED, FILEGROWTH = 10MB)
LOG ON -- Log files only fill 1 at a time. The second log file is just exposed here as a precautionary measure.
( NAME = N'ActorStage_TLOG_0', FILENAME = N'D:\OLTP_TLOG\ActorStage\ActorStage_TLOG_0.LDF' , SIZE = 102400KB , MAXSIZE = 204800KB , FILEGROWTH = 20480KB ),
( NAME = N'ActorStage_TLOG_1', FILENAME = N'D:\OLTP_TLOG\ActorStage\ActorStage_TLOG_1.LDF' , SIZE = 102400KB , MAXSIZE = 204800KB , FILEGROWTH = 20480KB )
GO
USE [ActorStage]
GO
CREATE SCHEMA [Metadata] AUTHORIZATION [dbo];
GO
CREATE SCHEMA [Hub] AUTHORIZATION [dbo];
GO
CREATE SCHEMA [Spoke] AUTHORIZATION [dbo];
GO
-- Force CS / AS because Actors are Sensitive about their Accents in each Case
ALTER DATABASE [ActorStage] COLLATE SQL_Latin1_General_CP1_CS_AS;
ALTER DATABASE [ActorStage] SET RECOVERY SIMPLE
--ALTER DATABASE [ActorStage] SET COMPATIBILITY_LEVEL = x
GO
ALTER DATABASE [ActorStage] SET ANSI_NULL_DEFAULT ON
ALTER DATABASE [ActorStage] SET ANSI_NULLS ON
ALTER DATABASE [ActorStage] SET ANSI_PADDING ON
ALTER DATABASE [ActorStage] SET ANSI_WARNINGS ON
ALTER DATABASE [ActorStage] SET QUOTED_IDENTIFIER OFF
GO
ALTER DATABASE [ActorStage] SET ARITHABORT ON
ALTER DATABASE [ActorStage] SET NUMERIC_ROUNDABORT ON
GO
ALTER DATABASE [ActorStage] SET AUTO_CLOSE OFF
ALTER DATABASE [ActorStage] SET AUTO_SHRINK OFF
ALTER DATABASE [ActorStage] SET AUTO_UPDATE_STATISTICS ON
GO
ALTER DATABASE [ActorStage] SET CURSOR_CLOSE_ON_COMMIT OFF
ALTER DATABASE [ActorStage] SET CURSOR_DEFAULT GLOBAL
ALTER DATABASE [ActorStage] SET CONCAT_NULL_YIELDS_NULL ON
GO
ALTER DATABASE [ActorStage] SET RECURSIVE_TRIGGERS OFF
ALTER DATABASE [ActorStage] SET DISABLE_BROKER
ALTER DATABASE [ActorStage] SET AUTO_UPDATE_STATISTICS_ASYNC OFF
ALTER DATABASE [ActorStage] SET DATE_CORRELATION_OPTIMIZATION OFF
ALTER DATABASE [ActorStage] SET TRUSTWORTHY OFF
ALTER DATABASE [ActorStage] SET ALLOW_SNAPSHOT_ISOLATION OFF
ALTER DATABASE [ActorStage] SET PARAMETERIZATION SIMPLE
ALTER DATABASE [ActorStage] SET READ_COMMITTED_SNAPSHOT OFF
ALTER DATABASE [ActorStage] SET HONOR_BROKER_PRIORITY OFF
GO
ALTER DATABASE [ActorStage] SET PAGE_VERIFY CHECKSUM
ALTER DATABASE [ActorStage] SET DB_CHAINING OFF
ALTER DATABASE [ActorStage] SET FILESTREAM( NON_TRANSACTED_ACCESS = OFF )
ALTER DATABASE [ActorStage] SET TARGET_RECOVERY_TIME = 0 SECONDS
GO
ALTER DATABASE [ActorStage] SET READ_WRITE
ALTER DATABASE [ActorStage] SET MULTI_USER
GO
IF (1 = FULLTEXTSERVICEPROPERTY('IsFullTextInstalled'))
BEGIN
EXECUTE [ActorStage].[dbo].[sp_fulltext_database] @action = 'enable'
END
GO
下一个：
/*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- Actors Stage
-------------------------------------------------
-- This Script Name: CreateTableIntervals.sql
-- Creator: @CubeSpark
-- Created: 2020 COVID Days
-- Purpose: Creates the Intervals table in the Spoke schema for SQL Pro
-- Compats: Tested on SQL Server 2019
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
DROP TABLE IF EXISTS Hub.Intervals;
PRINT 'Table Hub.Intervals will be created.';
GO
CREATE TABLE [Hub].[Intervals](
[Interval_PSK] INTEGER NOT NULL,
[Classifier] INTEGER NOT NULL,
[DatetimeIN] DATETIME NOT NULL,
[DatetimeOUT] DATETIME NOT NULL
CONSTRAINT [Intervals_PKC] PRIMARY KEY CLUSTERED -- index
(
[Interval_PSK], [Classifier], [DatetimeIN], [DatetimeOUT] ASC
) WITH (PAD_INDEX = OFF, STATISTICS_NORECOMPUTE = OFF, IGNORE_DUP_KEY = OFF, ALLOW_ROW_LOCKS = ON, ALLOW_PAGE_LOCKS = ON) ON [HUB]
) ON [HUB];
GO
IF OBJECT_ID (N'Hub.Intervals', N'U') IS NOT NULL
PRINT 'Table Hub.Intervals was successfully Created.';
GO
接下来，OLEDB OPENROWSET BULK PROVIDER 格式文件
<?xml version="1.0"?>
<BCPFORMAT xmlns="http://schemas.microsoft.com/sqlserver/2004/bulkload/format" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
<RECORD>
<FIELD ID="1" xsi:type="CharTerm" TERMINATOR="\t"/>
<FIELD ID="2" xsi:type="CharTerm" TERMINATOR="\t"/>
<FIELD ID="3" xsi:type="CharTerm" TERMINATOR="\t"/>
<FIELD ID="4" xsi:type="CharTerm" TERMINATOR="\t"/>
<FIELD ID="5" xsi:type="CharTerm" TERMINATOR="\r\n"/>
</RECORD>
<ROW>
<COLUMN SOURCE="1" NAME="Interval_PSK" xsi:type="SQLINT"/>
<COLUMN SOURCE="2" NAME="Classifier" xsi:type="SQLINT"/>
<COLUMN SOURCE="3" NAME="DateTimeIN" xsi:type="SQLDATETIME"/>
<COLUMN SOURCE="4" NAME="DateTimeOUT" xsi:type="SQLDATETIME"/>
</ROW>
</BCPFORMAT>
接下来，播种 SPROC 和错误 SPROC
/*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- Systems Deployment Management.
-------------------------------------------------
-- This Script Name: CreateGetErrorData.sql
-- Creator: @CubeSpark
-- Created: 2020 COVID Days
This procedure will obtain error data from Catch blocks.
-- Compats: Tested on SQL Server 2019
--ERROR_NUMBER() returns the number of the error.
--ERROR_SEVERITY() returns the severity.
--ERROR_STATE() returns the error state number.
--ERROR_PROCEDURE() returns the name of the stored procedure or trigger
--ERROR_LINE() returns the line number inside the routine that caused
--ERROR_MESSAGE() returns the complete text of the error message.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
DROP PROCEDURE IF EXISTS Meta.GetErrorData;
GO
IF OBJECT_ID(N'Meta.GetErrorData') IS NULL
BEGIN
PRINT 'Stored Procedure Meta.GetErrorData will be created.'
END;
GO
-- Create procedure to retrieve error information.
CREATE PROCEDURE Meta.GetErrorData
@InsertErrorRowOnly BIT = 0,
@SetMessageText NVARCHAR(200) = 'Calling SQL Script is not known'
AS
SET NOCOUNT ON;
IF @InsertErrorRowOnly = 0 -- use this if you want to just print the statements in the results window
BEGIN
SELECT
'UserName' = 'CubeSpark',
'ErrorNumber' = ERROR_NUMBER(),
'ErrorSeverity' = ERROR_SEVERITY(),
'ErrorState' = ERROR_STATE(),
'ErrorProcedure' = ISNULL(ERROR_PROCEDURE(), @SetMessageText),
'ErrorLine' = ERROR_LINE(),
'ErrorMessage' = ERROR_MESSAGE();
END;
IF @InsertErrorRowOnly = 1 -- use this if you want to register the error in the meta data table.
BEGIN
INSERT INTO [Meta].[ErrorData] (
[UserName],
[ErrorNumber],
[ErrorSeverity],
[ErrorState],
[ErrorProcedure],
[ErrorLine],
[ErrorMessage],
[ErrorDateTime])
SELECT
'UserName' = 'CubeSpark',
'ErrorNumber' = ERROR_NUMBER(),
'ErrorSeverity' = ERROR_SEVERITY(),
'ErrorState' = ERROR_STATE(),
'ErrorProcedure' = ISNULL(ERROR_PROCEDURE(), @SetMessageText),
'ErrorLine' = ERROR_LINE(),
'ErrorMessage' = ERROR_MESSAGE(),
'ErrorDateTime' = GETDATE();
END;
GO
IF OBJECT_ID(N'Meta.GetErrorData') IS NOT NULL
BEGIN
PRINT 'Stored Procedure Meta.GetErrorData has been created.'
END
接下来是播种和装载机。您需要 SQL Pro 帖子中的数据
/*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- Data Seed Spreading and Loading Script
-------------------------------------------------
-- SQL Server 2019 Version
-- This Script Name: CreateProcedureSeeding.sql
-- Creator: Mr. Grant Peace
-- Alterer: Mr. Getauf Thescaffolt
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
-- Use to make it USEFUL is your job
DROP PROCEDURE IF EXISTS [Meta].[Seeding];
PRINT 'Stored Procedure Meta.Seeding will be recreated'
GO
CREATE PROCEDURE [Meta].[Seeding]
@SchemaName NVARCHAR(20) = 'Hub',
@TableName NVARCHAR(100) = 'Intervals',
@StartingRow INTEGER = 1,
@CodePage INTEGER = 1252,
@UNC_QualifiedErrorFilePath NVARCHAR(200) = '\\YOURTHING\OutputFiles\Errors.txt',
@UNC_QualifiedSeedFilePath NVARCHAR(200) = '\\YOURTHING\InputFiles\Intervals.txt',
@UNC_QualifiedFormatFilePath NVARCHAR(200) = '\\YOURTHING\OutputFiles\Intervals.xml',
@PrintOnly BIT = 1,
@WithTruncate BIT = 0,
@InsertErrorRowOnly BIT = 0,
@Message_OUT NVARCHAR(MAX) = 'Test' OUTPUT,
@ErrorMessage_OUT NVARCHAR(MAX) = 'Test' OUTPUT,
@RowCount_OUT INTEGER = 0 OUTPUT
AS
SET NOCOUNT ON
DECLARE
@Executable NVARCHAR(MAX),
@RowCount INTEGER,
@crlf NCHAR(2) = CHAR(13) + CHAR(10),
@Parameters NVARCHAR(100),
@SetMessageText NVARCHAR(1024) = 'No Errors Handled';
SET @Parameters = N'@RowCount_EX INTEGER OUTPUT'
SET @RowCount_OUT = 0;
SET @Message_OUT = 'Success.';
SET @ErrorMessage_OUT = 'No errors.';
-- No table exists, so then you can do a SELECT INTO.
BEGIN TRY
IF @WithTruncate = 1
BEGIN
SET @Executable =
'IF OBJECT_ID (N'
+ '''' + '[' + @SchemaName + '].['
+ @TableName + ']' + '''' + ', N' + '''' + 'U' + '''' + ') IS NOT NULL '
+ @crlf
+ 'BEGIN '
+ @crlf
+ ' TRUNCATE TABLE [' + @SchemaName + '].[' + @TableName + '];'
+ @crlf
+ ' INSERT INTO [' + @SchemaName + '].[' + @TableName + '] WITH (KEEPDEFAULTS)'
+ @crlf
+ ' SELECT SourceLoad.* FROM OPENROWSET(BULK ' + '''' + @UNC_QualifiedSeedFilePath + '''' + ', '
+ @crlf
+ ' FORMATFILE = ' + '''' + @UNC_QualifiedFormatFilePath + '''' + ', '
+ @crlf
+ ' ERRORFILE = ' + '''' + @UNC_QualifiedErrorFilePath + '''' + ', '
+ @crlf
+ ' FIRSTROW = ' + CAST(@StartingRow AS NVARCHAR(10)) + ', ' --2,'
+ @crlf
+ ' CODEPAGE = ' + '''' + CAST(@CodePage AS NVARCHAR(6)) + '''' + ') AS SourceLoad'
+ @crlf
+ 'END;'
+ @crlf
+ 'SET @RowCount_EX = @@ROWCOUNT;'
END
ELSE
BEGIN
SET @Executable =
'IF OBJECT_ID (N'
+ '''' + '[' + @SchemaName + '].['
+ @TableName + ']' + '''' + ', N' + '''' + 'U' + '''' + ') IS NOT NULL '
+ @crlf
+ 'BEGIN '
+ @crlf
+ ' INSERT INTO [' + @SchemaName + '].[' + @TableName + '] WITH (KEEPDEFAULTS)'
+ @crlf
+ ' SELECT SourceLoad.* FROM OPENROWSET(BULK ' + '''' + @UNC_QualifiedSeedFilePath + '''' + ', '
+ @crlf
+ ' FORMATFILE = ' + '''' + @UNC_QualifiedFormatFilePath + '''' + ', '
+ @crlf
+ ' ERRORFILE = ' + '''' + @UNC_QualifiedErrorFilePath + '''' + ', '
+ @crlf
+ ' FIRSTROW = ' + CAST(@StartingRow AS NVARCHAR(10)) + ', ' --2,'
+ @crlf
+ ' CODEPAGE = ' + '''' + CAST(@CodePage AS NVARCHAR(6)) + '''' + ') AS SourceLoad'
+ @crlf
--+ @crlf
--+ ' FIRSTROW = 2,'
--+ ' CODEPAGE = ' + '''' + 'RAW' + '''' + ') AS SourceLoad'
+ @crlf
+ 'END;'
+ @crlf
+ 'SET @RowCount_EX = @@ROWCOUNT;'
END;
IF @PrintOnly = 1 PRINT @Executable -- SSIS Suppresses PRINT because its not STDIO.
IF @PrintOnly = 0 EXECUTE sp_executesql @Executable, @Parameters, @RowCount_EX = @RowCount_OUT OUTPUT;
IF @PrintOnly = 0 AND @RowCount_OUT > 0
BEGIN
PRINT 'Seeding of [' + @SchemaName + '].[' + @TableName + '] is complete, with an affected row count of: ' + CAST(@RowCount_OUT AS VARCHAR(10));
END
IF @PrintOnly = 0 AND @RowCount_OUT = 0
BEGIN
PRINT 'Table [' + @SchemaName + '].[' + @TableName + '] does not exist. Create the table first before loading.'
END
END TRY
BEGIN CATCH
-- This will only register faults in the dynamic SQL.
PRINT 'Inside Catch'
SET @SetMessageText = N'Caller: ' + OBJECT_SCHEMA_NAME(@@PROCID) + '.' + OBJECT_NAME(@@PROCID);
EXECUTE Meta.GetErrorData
@SetMessageText = @SetMessageText,
@InsertErrorRowOnly = @InsertErrorRowOnly;
SET @Message_OUT = 'Failure';
SET @ErrorMessage_OUT = 'Error Number #: '
+ CONVERT(NVARCHAR(15), ERROR_NUMBER())
+ '.' + @crlf + 'Error Message: '
+ ERROR_MESSAGE()
+ @crlf + 'Module line #: '
+ CONVERT(NVARCHAR(15), ERROR_LINE())
+ @crlf + 'Error Severity: '
+ CONVERT(NVARCHAR(10), ERROR_SEVERITY())
+ @crlf + 'Error State: '
+ CONVERT(NVARCHAR(10), ERROR_STATE())
PRINT @ErrorMessage_OUT;
END CATCH;
GO
IF OBJECT_ID(N'Meta.Seeding') IS NOT NULL
BEGIN
PRINT 'Stored Procedure Meta.Seeding was recreated.'
END
GO
Next: Data Loader. Pay careful attention to the folder and the parameters.
/*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-- Stack Exchange
-- Data Loading SPROC
-------------------------------------------------
https://dba.stackexchange.com/questions/275487/2019-enterprise-nolock-without-nolock-page-latch-time-out-on-tempdb-and-final/280977#280977
-- SQL Server 2019 Version
-- This Script Name: LoadIntervals.sql
-- Creator: Mr. Getauf Thescaffolt
-- Created: 2018-03-31 09:00:00.000
-- Altered: 2019-07-28 09:00:00.000
-- Alterer: Mr. Getauf Thescaffolt
-- Outputs: Data to table
-- Purpose:
-- Package:
-- Folders:
Requires Format File, UNC Paths, and TAB Text as specified by SQL Pro
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
USE [Quantum]
GO
DECLARE @RC INTEGER,
@SchemaName NVARCHAR(20),
@TableName NVARCHAR(100),
@UNC_QualifiedErrorFilePath NVARCHAR(200),
@UNC_QualifiedSeedFilePath NVARCHAR(200),
@UNC_QualifiedFormatFilePath NVARCHAR(200),
@PrintOnly BIT,
@WithTruncate BIT,
@InsertErrorRowOnly BIT,
@Message_OUT NVARCHAR(MAX),
@ErrorMessage_OUT NVARCHAR(MAX),
@RowCount_OUT INTEGER;
EXECUTE [Meta].[Seeding]
@SchemaName = [Hub],
@TableName = [Intervals],
@CodePage = 1252,
@StartingRow = 1,
@UNC_QualifiedErrorFilePath = '\\YOURTHING\OutputFiles\Errors.txt',
@UNC_QualifiedSeedFilePath = '\\YOURTHING\InputFiles\Intervals.tx