Posted on

Get data to lab report

In this post we continue with lab series. I fill prepare structures with tested data. To collect testing results I will use Extended Events, Extended Events utility from this post.

In post we will demonstrate very simple demonstration of clustered index impact on query performance. One table will be created with clustered index and one table stay as heap. After filling tables with data, we start Extended Event Session and run two queries against created tables. When queries finish, we collect the data to lab structure we created in previous post and stop and clean sessions.

Let’s prepare lab structures first.

CREATE TABLE dbo.TestTableData
(id INT )
 
CREATE TABLE dbo.TestTableDatClustered
(id INT PRIMARY KEY CLUSTERED
)
 
INSERT INTO dbo.TestTableData
SELECT TOP 1000000 ROW_NUMBER() OVER (ORDER BY (SELECT 1)) rn 
FROM sys.objects a
JOIN sys.objects b ON 1=1
JOIN sys.objects c ON 1=1
 
INSERT INTO dbo.TestTableDatClustered
SELECT TOP 1000000 ROW_NUMBER() OVER (ORDER BY (SELECT 1)) rn 
FROM sys.objects a
JOIN sys.objects b ON 1=1
JOIN sys.objects c ON 1=1

Clean cache and get SPID.

SELECT @@SPID
DBCC FREEPROCCACHE
CHECKPOINT
DBCC DROPCLEANBUFFERS

Open new query window and start session with the utility.Extended events monitoring utility

Create and start session.

/*Lets start event session*/ EXEC mnt.MNT_SessionHandler 'CreateSession;StartSession','Test','QueryPlan;IOStat','<Predicate><DatabaseID>5</DatabaseID></Predicate>' /*Lets start events*/

Run test queries.

SELECT * FROM dbo.TestTableData WHERE Id = 1000
SELECT * FROM dbo.TestTableDatClustered WHERE Id = 1000

Get data from Session to temporary table.

EXEC mnt.MNT_SessionHandler 'GetDataFromSession' ,'Test','IOStat;QueryPlan' /*Run query you would like to analyze*/

Let’s clean up the extended event session.

/*Stop session*/ EXEC mnt.MNT_SessionHandler 'EndSession', 'Test' /*clean session*/ EXEC mnt.MNT_SessionHandler 'CleanSession', 'Test' /*Or all in one raw*/ EXEC mnt.MNT_SessionHandler 'CleanSession;EndSession', 'Test' /*Get data to analyze,lets make your own views, predicates... */ SELECT * FROM ##Test

In the script bellow the data from extended events are stored to Lab structures.

 
INSERT INTO dbo.LabTable
( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT CPUTime/ 1000,'CPUTime (sec)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText = ' SELECT * FROM dbo.TestTableData WHERE Id = 1000'
AND [Type] = 'QueryPlan'
INSERT INTO dbo.LabTable
( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT Duration/ 1000,'Duration (sec)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText = ' SELECT * FROM dbo.TestTableData WHERE Id = 1000'
AND [Type] = 'QueryPlan'
INSERT INTO dbo.LabTable
( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT PhysicalReads,'PhysicalReads (pages)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText = ' SELECT * FROM dbo.TestTableData WHERE Id = 1000' AND [Type] = 'IOSql'
INSERT INTO dbo.LabTable
( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT LogicalReads,'LogicalReads (pages)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText = ' SELECT * FROM dbo.TestTableData WHERE Id = 1000' AND [Type] = 'IOSql'
 
INSERT INTO dbo.LabTable( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT CPUTime/ 1000,'CPUTime (sec)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText LIKE '%dbo.TestTableDatClustered%' AND [Type] = 'QueryPlan'
INSERT INTO dbo.LabTable ( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT Duration/ 1000,'Duration (sec)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText LIKE '%dbo.TestTableDatClustered%' AND [Type] = 'QueryPlan'

INSERT INTO dbo.LabTable ( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT PhysicalReads,'PhysicalReads (pages)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText LIKE '%dbo.TestTableDatClustered%' AND [Type] = 'IOSql'

INSERT INTO dbo.LabTable ( [CounterValue],[CounterName],[TaskOwnerName], [WorkoutName] )
SELECT LogicalReads,'LogicalReads (pages)', SqlText,'ClusteredIndexTest'
FROM dbo.##Test WHERE SqlText LIKE '%dbo.TestTableDatClustered%' AND [Type] = 'IOSql'

In next post we visualise the data in SSRS chart, so stay tuned.

Posted on

Get data to lab report

I started with series of post, I would like to show how to prepare simple testing environment, you can get the data when you test or compare queries performance for example, and you would like to save results and visualise them.

In this post we create simple table structure and SQL procedure which will be used in next post as Chart dataset.

I would like to save following information’s:

  • Test identification – identification of test – for example testing clustered vs nanocluster index, sorting group vs hash group, etc.
  • Task name – identification of task we made in test – for example task1 = table with clustered index, task2 heap table
  • Counter name (for example number of reads, writes, rows, etc.)
  • Counter value – measured results of tests

As you can see the structure is very simple, I believe we could design much sophisticated solution but for our purposes will be enough. Next time we can extend our solution.

Here is the final structure:

SET
ANSI_NULLS
ON
GO
SET
QUOTED_IDENTIFIER
ON
GO
 
CREATE TABLE [dbo].[LabTable](
    [WorkoutName] [sysname] NOT NULL,
    [TaskOwnerName] [sysname] NOT NULL,
    [CounterName] [sysname] NOT NULL,
    [CounterValue] [sql_variant] NULL,
    [InsertDateTime] [datetime] NULL
)
ON [PRIMARY]
GO
 
ALTER
TABLE [dbo].[LabTable] ADD
CONSTRAINT [df_InsertDateTime] DEFAULT (getdate())
FOR [InsertDateTime]
GO

Stored procedure returning data to dataset:

SET
ANSI_NULLS
ON
GO
SET QUOTED_IDENTIFIER ON
GO
 
CREATE PROCEDURE [dbo].[SSRSGetDatset]
@WorkoutName SYSNAME
AS
BEGIN
SELECT WorkoutName,
     TaskOwnerName,
     CounterName,
     TRY_CONVERT(BIGINT,TRY_CONVERT(DATETIME, CounterValue))
AS CounterValue,
     InsertDateTime
FROM dbo.LabTable
WHERE WorkoutName = @WorkoutName
END
GO

In next post we fill structures with tested data and create graph to visualise our results. Stay tuned.

Posted on

Extended events advanced

 I would like to describe useful queries for extended events in more detail in this post Extended events get description. I will follow my previous post where I put few basic queries to get some extended events info. 

Extended events are triggers occurring in SQL system. Each extended event has its own set of properties -data you get from result. Extended events are collected in packages, events can be combined from different packages. List of packages defined in SQL server instance  

SELECT * FROM sys.dm_xe_packages

 Beside of that each event can fire other actions synchronously int the same thread within event was invoked. Action extend extended events with own data in response.

For example, if you wold like to get sql text of a query it is defined by an action.

SELECT *
  FROM sys.dm_xe_objects o 
 WHERE o.object_type='action'

To filter event data, you can define Predicates. There is a list of objects you can use as predicate source. 

 SELECT *
 FROM sys.dm_xe_objects o 
WHERE o.object_type='pred_source'

List of operators 

SELECT *
  FROM sys.dm_xe_objects o 
 WHERE o.object_type='pred_compare'

 To start collecting data you must define session object which is set of actions, event data, predicates and target object.

 Target object collects output event data by synchronous or asynchronous way. Data can be collected for example to buffer, file, Windows Events, etc.

SELECT * FROM sys.dm_xe_objects o WHERE o.object_type='target' 

 User – friendly description of internal key words which could be returned from events data 

SELECT *
  FROM sys.dm_xe_objects o 
 WHERE o.object_type='map' 

  Relationship between events implemented in created sessions and actions 

SELECT * FROM sys.dm_xe_session_event_actions

 Data types you can expect from data collection. 

SELECT *
  FROM sys.dm_xe_objects o 
 WHERE o.object_type='type'

 Next time I would like to share some solutions implemented using extended events. Stay tuned.