?
Solved

Need help with a qry's efficiency

Posted on 2009-05-15
13
Medium Priority
?
214 Views
Last Modified: 2012-05-07
Friends,

I have a qry (attached) that is taking forever to run.   Can anyone look to see where the problem lies?  It really runs into trouble with the Order By.  If I take out the group by and the order by statement, it works great, it just doesn't order it right (obviously).

Thanks in advance!

Best Regards,
Eric
Select
t.TrackOrder,
RTRIM(r.No) as CarNumber,
p.ResultItemID,
p.PassingTime,
p.ElapsedTime,
p.PassingID,
p.DecoderID,
p.LeaderLap, 
p.LapCount 
From Passings p with (nolock) inner join Results r with (nolock)
ON p.RunID=r.RunID
AND p.ResultItemID=r.ResultID
Inner Join Timelines t 
on t.DecoderID=p.DecoderID
AND t.LoopID=p.LoopID
Where 
p.Pit = 0
And 
passingtime in (SELECT max(i.passingtime) FROM passings i with (NOLOCK) WHERE
    i.ResultItemID = p.ResultItemID
    AND i.passingtime >=0
    AND i.lapcount>0
    AND i.Deleted=0
    AND p.Deleted=0
    AND i.RunID=6
	AND p.RunID=6
    AND i.DecoderID<> 0)--(Select Top 1 DecoderID From Timelines where IsLapCounter=1))
And p.RunID=6
And r.RunID=6
Group by p.ResultItemID, 
p.PassingTime, 
p.ElapsedTime, 
p.PassingID, 
p.DecoderID, 
p.LapCount, 
p.LeaderLap, 
t.TrackOrder,
r.No, 
r.RaceRank
Order By t.TrackOrder Desc, p.ElapsedTIme

Open in new window

0
Comment
Question by:indy500fan
13 Comments
 
LVL 6

Expert Comment

by:bull_rider
ID: 24397616
You have missed one more WITH (NOLOCK) in this statement

Inner Join Timelines t

Can you add and check?
0
 
LVL 75

Expert Comment

by:Aneesh Retnakaran
ID: 24397619
pls post the table structue including the indexes
0
 
LVL 9

Expert Comment

by:cdaly33
ID: 24397630
You have a couple options.  You could try putting an index on the t.TrackOrder and p.ElapsedTIme fields which may or may not help depending on the amount of data.  Or you can try to find a way to sort it after the data has been retrieved such as in the program utilizing the data.  Or you if this is in a stored procedure you could create a temp table variable matching the structure that is returned from the statement and then select out of that temp table variable with the appropriate sorting applied then.
0
Veeam and MySQL: How to Perform Backup & Recovery

MySQL and the MariaDB variant are among the most used databases in Linux environments, and many critical applications support their data on them. Watch this recorded webinar to find out how Veeam Backup & Replication allows you to get consistent backups of MySQL databases.

 

Author Comment

by:indy500fan
ID: 24397695
ttached is the passings table structure.  By the way, the data inside the Timelines table doesn't change.

SET QUOTED_IDENTIFIER ON
GO
CREATE TABLE [dbo].[Passings](
	[RunID] [int] NOT NULL,
	[PassingID] [int] NOT NULL,
	[FlagState] [int] NOT NULL,
	[DecoderID] [int] NOT NULL,
	[TranNr] [int] NOT NULL,
	[PassingTime] [int] NOT NULL,
	[PhotocellTime] [int] NOT NULL,
	[ElapsedTime] [int] NOT NULL,
	[Hits] [int] NOT NULL,
	[Strength] [int] NOT NULL,
	[LowBatt] [bit] NOT NULL,
	[LoopID] [int] NOT NULL,
	[Pit] [bit] NOT NULL,
	[Deleted] [bit] NOT NULL,
	[Manual] [bit] NOT NULL,
	[Photocell] [bit] NOT NULL,
	[PhotocellState] [int] NOT NULL,
	[Flag] [bit] NOT NULL,
	[LeaderLap] [int] NOT NULL,
	[LapCount] [int] NOT NULL,
	[Laptime] [int] NOT NULL,
	[LaptimeViolation] [bit] NOT NULL,
	[Finish] [bit] NOT NULL,
	[ResultItemID] [int] NOT NULL,
	[EventID] [int] NOT NULL DEFAULT ((1000)),
	[SeasonID] [int] NOT NULL DEFAULT ((1001))
) ON [PRIMARY]
 
/****** Object:  Index [PASSINGID_INDEX]    Script Date: 05/15/2009 12:45:05 ******/
CREATE NONCLUSTERED INDEX [PASSINGID_INDEX] ON [dbo].[Passings] 
(
	[PassingID] DESC
)WITH (PAD_INDEX  = OFF, STATISTICS_NORECOMPUTE  = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS  = ON, ALLOW_PAGE_LOCKS  = ON) ON [PRIMARY]
 
/****** Object:  Table [dbo].[Results]    Script Date: 05/15/2009 12:45:56 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
SET ANSI_PADDING ON
GO
CREATE TABLE [dbo].[Results](
	[RunID] [int] NOT NULL,
	[ResultID] [int] NOT NULL,
	[TranNr] [int] NOT NULL,
	[Deleted] [bit] NOT NULL,
	[ExtraTotaltime] [int] NOT NULL,
	[ExtraLapcount] [int] NOT NULL,
	[ExtraBestLaptime] [int] NOT NULL,
	[OverallBestLaptime] [int] NOT NULL,
	[ResultStatus] [int] NOT NULL,
	[StartPosition] [int] NOT NULL,
	[Comment] [char](255) NOT NULL,
	[QualifyRank] [int] NOT NULL,
	[RaceRank] [int] NOT NULL,
	[BestLaptime] [int] NOT NULL,
	[BestLap] [int] NOT NULL,
	[BestSpeed] [float] NOT NULL,
	[SecondBestLaptime] [int] NOT NULL,
	[SecondBestLap] [int] NOT NULL,
	[SecondBestSpeed] [float] NOT NULL,
	[LastLaptime] [int] NOT NULL,
	[LastSpeed] [float] NOT NULL,
	[Laps] [int] NOT NULL,
	[Totaltime] [int] NOT NULL,
	[DiffQual] [char](70) NOT NULL,
	[GapQual] [char](70) NOT NULL,
	[DiffRace] [char](70) NOT NULL,
	[GapRace] [char](70) NOT NULL,
	[CompetitorID] [int] NOT NULL,
	[RaceMarker] [int] NOT NULL,
	[QualMarker] [int] NOT NULL,
	[AvgSpeed] [float] NOT NULL,
	[OnTrack] [int] NOT NULL,
	[PitStops] [int] NOT NULL,
	[LastPitLap] [int] NOT NULL,
	[SincePit] [int] NOT NULL,
	[SincePitLap] [int] NOT NULL,
	[FlagStatus] [char](255) NOT NULL,
	[No] [char](70) NOT NULL,
	[FirstName] [char](70) NOT NULL,
	[LastName] [char](70) NOT NULL,
	[Class] [char](70) NOT NULL,
	[DriverID] [char](70) NOT NULL,
	[TrpndrCtrl#] [char](70) NOT NULL,
	[Equipment] [char](70) NOT NULL,
	[License] [char](70) NOT NULL,
	[Team] [char](70) NOT NULL,
	[TeamID] [char](70) NOT NULL,
	[Engine] [char](70) NOT NULL,
	[EntrantID] [char](70) NOT NULL,
	[EventID] [int] NOT NULL DEFAULT ((1000)),
	[SeasonID] [int] NOT NULL DEFAULT ((1001))
 
) ON [PRIMARY]
 
GO
SET ANSI_PADDING OFF
 
GO
/****** Object:  Index [RESULT_RUNID_INDEX]    Script Date: 05/15/2009 12:47:15 ******/
CREATE NONCLUSTERED INDEX [RESULT_RUNID_INDEX] ON [dbo].[Results] 
(
	[RunID] DESC
)WITH (PAD_INDEX  = OFF, STATISTICS_NORECOMPUTE  = OFF, SORT_IN_TEMPDB = OFF, IGNORE_DUP_KEY = OFF, DROP_EXISTING = OFF, ONLINE = OFF, ALLOW_ROW_LOCKS  = ON, ALLOW_PAGE_LOCKS  = ON) ON [PRIMARY]
 
 
/****** Object:  Table [dbo].[Timelines]    Script Date: 05/15/2009 12:48:11 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
SET ANSI_PADDING ON
GO
CREATE TABLE [dbo].[Timelines](
	[m_UniqueID] [int] NOT NULL,
	[TrackID] [int] NOT NULL,
	[Name] [char](70) NOT NULL,
	[DecoderID] [int] NOT NULL,
	[LoopID] [int] NOT NULL,
	[TrackOrder] [int] NOT NULL,
	[DistanceFromSF] [float] NOT NULL,
	[IsSendToFeed] [bit] NOT NULL,
	[IsLapCounter] [bit] NOT NULL,
	[IsInPit] [bit] NOT NULL,
	[IsPitIn] [bit] NOT NULL,
	[IsPitOut] [bit] NOT NULL,
	[IsFirstAfterPitOut] [bit] NOT NULL,
	[IsBlendline] [bit] NOT NULL,
	[EventID] [int] NOT NULL DEFAULT ((1000)),
	[SeasonID] [int] NOT NULL DEFAULT ((1001))
) ON [PRIMARY]
 
GO
SET ANSI_PADDING OFF

Open in new window

0
 
LVL 75

Expert Comment

by:Aneesh Retnakaran
ID: 24397870
Question?
Do you really need a GROUP BY part ?
I am not sure about the size of and values on the timelines table, can u create the following index on that table and try it..
if possibgle post the execution plan too
create index idx_Timelines_DecoderID_LoopID_TrackOrder ON TimeLines (DecoderID) INCLUDE (LoopID,TrackOrder)  
0
 

Author Comment

by:indy500fan
ID: 24397947
aneeshattingal,

I have tried it with the index.  No change.  Execution Plan attached as image.
ExecutionPlan.bmp
0
 
LVL 75

Expert Comment

by:Aneesh Retnakaran
ID: 24398005
You need to create indexes on the RUNID's (both the tables)
Wgy dont you have a Primary key / Clustured indexes on those tables /
0
 

Author Comment

by:indy500fan
ID: 24398052
These tables are created by 3rd pary software  I'm not very good with DB's, and I'm weary about creating indexes.  I have very little sql experience.

Do you have suggestions as to creating the indexes?  Which fields on each table or do I create a new field and seed it?
0
 
LVL 75

Accepted Solution

by:
Aneesh Retnakaran earned 2000 total points
ID: 24398148
create index idx_Results_ResultID_RunID ON Results (RunID) INCLUDE (ResultID)  
create index idx_Passings_ResultItemID_RunID ON Passings (RunID) INCLUDE (ResultItemID)
0
 

Author Comment

by:indy500fan
ID: 24398208
Interesting.  I ran it and it was faster against today's dataset, but I need to try one more test tomorrow morning with that dataset.

Thank you for your help so far!

Regards,
Eric
0
 
LVL 75

Expert Comment

by:Aneesh Retnakaran
ID: 24398233
the query can be even faster with a primary key &clustured index, as a rule you should make sure that you have a primary key for all the tables
0
 

Author Comment

by:indy500fan
ID: 24398280
I may try that!  I really think we are on the right path!
0
 

Author Comment

by:indy500fan
ID: 24402936
The indexes did the trick!  Thank you!
0

Featured Post

What does it mean to be "Always On"?

Is your cloud always on? With an Always On cloud you won't have to worry about downtime for maintenance or software application code updates, ensuring that your bottom line isn't affected.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Recently, when I was asked to create a new SQL 2005 cluster, Microsoft released a new service pack for MS SQL 2005 what is Service Pack 3. When I finished the installation of MS SQL 2005 I found myself troubled why the installation of SP3 failed …
by Mark Wills PIVOT is a great facility and solves many an EAV (Entity - Attribute - Value) type transformation where we need the information held as data within a column to become columns in their own right. Now, in some cases that is relatively…
Integration Management Part 2
Are you ready to place your question in front of subject-matter experts for more timely responses? With the release of Priority Question, Premium Members, Team Accounts and Qualified Experts can now identify the emergent level of their issue, signal…
Suggested Courses

755 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question