Inside the SQL Server Query Optimizer - part 6 What is the memo?

Hi Guys,

Where did we get to last time?

Last time we had talked about the cost based optimization process and the transformation rules that applied by the optimizer were looking for a better performance plan.

The memo

So the question is: Where our alternative plans are stored?
They are stored in a component called memo.

You can find more info about this structure in the “Cascades general optimization framework” developed by Goetz Graefe.

Today I would show you only how the memo work, nothing more!
We should enable some trace flag in order to display the information we need.
  • Through the trace flag 8608 we can show the initial Memo Structure,
  • Through the trace flag 8615 we can show the final Memo Structure,

To better explain the process we take as example our Query used in the previous posts.
       
SELECT
  ANNDOC, NUMDOC, SUM(R.QTA1) AS SUM_QTA
FROM ORDTES T
  JOIN ORDRIG R ON R.IDORDTES = T.ID
WHERE
  ANNDOC = 2018 
GROUP BY ANNDOC, NUMDOC
ORDER BY ANNDOC, NUMDOC
OPTION ( RECOMPILE,
          querytraceon 2363,
          querytraceon 3604,
          querytraceon 8606,  -- race flag 8606 will display additional logical trees used during the optimization process
          querytraceon 8612,  -- Add Extra Info to the Trees Output
          querytraceon 8619,  -- Show Applied Transformation Rules
          querytraceon 8620,  -- 8620 = Add memo arguments to trace flag 8619
          querytraceon 8608,  -- Shows the initial Memo structure
          querytraceon 8615,  -- Display the final memo structure
          querytraceon 8621)  -- 8621 = Rule with resulting tree

As seen in the previous post our query have this logical tree:





Through the flag 8608 we can take a look to the initial Memo Structure.

You can note, at this point, that the logical tree is simply copied into the memo.
One group per logical node.

       

 ****************************************
--- Initial Memo Structure ---

Root Group 14: Card=1635.67 (Max=8.4478e+010, Min=0)

   0 LogOp_GbAgg 9 13 (Distance = 0)


Group 13:
   0 AncOp_PrjList  12 (Distance = 0)


Group 12:
   0 AncOp_PrjEl  11 (Distance = 0)


Group 11:
   0 ScaOp_AggFunc  10 (Distance = 0)


Group 10:
   0 ScaOp_Identifier  (Distance = 0)


Group 9: Card=66408.4 (Max=8.4478e+010, Min=0)

   0 LogOp_Join 4 5 8 (Distance = 0)


Group 8:
   0 ScaOp_Comp  6 7 (Distance = 0)


Group 7:
   0 ScaOp_Identifier  (Distance = 0)


Group 6:
   0 ScaOp_Identifier  (Distance = 0)



Group 5: Card=566688 (Max=623357, Min=0)

   0 LogOp_Get (Distance = 0)


Group 4: Card=1641 (Max=135521, Min=0)

   0 LogOp_Select 0 3 (Distance = 0)


Group 3:
   0 ScaOp_Comp  1 2 (Distance = 0)


Group 2:
   0 ScaOp_Const  (Distance = 0)


Group 1:
   0 ScaOp_Identifier  (Distance = 0)


Group 0: Card=123201 (Max=135521, Min=0)

   0 LogOp_Get (Distance = 0)

Now at the end of the optimization procedure (so after the Search 0, Search 1 and Search 2 phases of witch we speak in detail in the next and last post of this series) we will have in the memo all the optimization applied.

Here our final Memo Structure after the Search optimization phases:

Remember: Each phases can add a new group or an existing group may be expanded to include other logical and physical alternative. 

       

--- Final Memo Structure ---

Group 32: Card=566688 (Max=623357, Min=0)

   0 LogOp_Spool 5 (Distance = 1)



Group 31: Card=46607.6 (Max=623357, Min=0)

   0 LogOp_Spool 22 (Distance = 3)



Group 30: Card=1 (Max=135521, Min=0)

   0 LogOp_SelectRes 29 3 (Distance = 4)



Group 29: Card=1 (Max=135521, Min=0)

   0 LogOp_SelectIdx 28 8 (Distance = 4)



Group 28: Card=123201 (Max=135521, Min=0)

   0 LogOp_GetIdx (Distance = 4)



Group 27: Card=1 (Max=135521, Min=0)

   1 PhyOp_Range 5 ASC  26.0  Cost(RowGoal 0,ReW 0.0512361,ReB 46606.6,Dist 46607.6,Total 46607.6)= 7.37882 (Distance = 5)


   0 LogOp_SelectIdx 25 26 (Distance = 4)



Group 26:
   0 ScaOp_Logical  8.0 3.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 7 (Distance = 4)



Group 25: Card=123201 (Max=135521, Min=0)

   0 LogOp_GetIdx (Distance = 4)



Group 24: Card=566688 (Max=623357, Min=0)

   1 PhyOp_Range 1 ASC   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 26.8615 (Distance = 2)


   0 LogOp_GetIdx (Distance = 1)



Group 23: Card=1641 (Max=135521, Min=0)

   3 PhyOp_HashJoinx_jtInner 4.1 22.1 8.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 32.1464 (Distance = 3)


   2 PhyOp_Sort 23.3  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 32.1851 (Distance = 0)


   1 LogOp_Join 22 4 8 (Distance = 3)


   0 LogOp_Join 4 22 8 (Distance = 2)



Group 22: Card=46607.6 (Max=623357, Min=0)

   6 PhyOp_Sort 22.1  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 35.1172 (Distance = 0)


   1 PhyOp_HashGbAgg 5.2 16.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 31.7882 (Distance = 3)


   0 LogOp_GbAgg 5 16 (Distance = 2)



Group 21:
   0 AncOp_PrjList  20.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 1)



Group 20:
   0 AncOp_PrjEl  19.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 1)



Group 19:
   0 ScaOp_AggFunc  18.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 2 (Distance = 1)



Group 18:
   0 ScaOp_Identifier   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 1)



Group 17: Card=6542.61 (Max=8.4478e+010, Min=0)

   2 PhyOp_HashGbAgg 9.2 16.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 31.1182 (Distance = 2)


   1 PhyOp_Sort 17.2  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 31.3337 (Distance = 0)


   0 LogOp_GbAgg 9 16 (Distance = 1)



Group 16:
   0 AncOp_PrjList  15.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 1)



Group 15:
   0 AncOp_PrjEl  11.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 1)



Root Group 14: Card=1635.67 (Max=8.4478e+010, Min=0)

   11 PhyOp_HashGbAgg 9.2 13.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 31.0412 (Distance = 1)


   9 PhyOp_Sort 14.11  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 31.0798 (Distance = 0)


   2 LogOp_GbAgg 23 21 (Distance = 2)


   1 LogOp_GbAgg 17 21 (Distance = 1)


   0 LogOp_GbAgg 9 13 (Distance = 0)



Group 13:
   0 AncOp_PrjList  12.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 0)



Group 12:
   0 AncOp_PrjEl  11.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0 (Distance = 0)



Group 11:
   0 ScaOp_AggFunc  10.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 2 (Distance = 0)



Group 10:
   0 ScaOp_Identifier   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 0)



Group 9: Card=66408.4 (Max=8.4478e+010, Min=0)

   8 PhyOp_Sort 9.2  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 35.5838 (Distance = 0)


   2 PhyOp_HashJoinx_jtInner 4.1 5.2 8.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 30.6896 (Distance = 1)


   1 LogOp_Join 5 4 8 (Distance = 1)


   0 LogOp_Join 4 5 8 (Distance = 0)



Group 8:
   0 ScaOp_Comp  6.0 7.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 3 (Distance = 0)



Group 7:
   0 ScaOp_Identifier   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 0)



Group 6:
   0 ScaOp_Identifier   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 0)



Group 5: Card=566688 (Max=623357, Min=0)

   6 PhyOp_Sort 5.2  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 76.5855 (Distance = 0)


   2 PhyOp_Range 1 ASC   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 26.8615 (Distance = 1)


   0 LogOp_Get (Distance = 0)



Group 4: Card=1641 (Max=135521, Min=0)

   1 PhyOp_Range 5 ASC  3.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 0.00675377 (Distance = 1)


   0 LogOp_Select 0 3 (Distance = 0)



Group 3:
   0 ScaOp_Comp  1.0 2.0  Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 3 (Distance = 0)



Group 2:
   0 ScaOp_Const   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 0)



Group 1:
   0 ScaOp_Identifier   Cost(RowGoal 0,ReW 0,ReB 0,Dist 0,Total 0)= 1 (Distance = 0)



Group 0: Card=123201 (Max=135521, Min=0)

   0 LogOp_Get (Distance = 0)

That's all for today.
I hope you enjoyed this post! if so wait for the last post of this series!
Have an amazing week!
See you soon!


Luca Biondi @ SQLServerPerformance blog!

Comments

Post a Comment

I Post più popolari

SQL Server, execution plan and the lazy spool (clearly explained)

Image
Hi Guys, Welcome back! In one of the last posts we talked about Table Spool . We had explained what Spool are for by telling the somewhat curious story of the halloween problem . Here we specifically talked about a type of spool called Eager Spool   You can look to the execution plan below where the eager Spool is used to avoid the halloween problem in classic query of the "salary increase".. i suggest Well, if the story of the eager spool intrigued you, today instead we will speak about another type of Spool, the Lazy Spool . The Lazy spool Let's take a small step back. So, what is a Spool operation?  As we have already understood a spool operation is simply a temporary storage of data . Data are read from a source table and stored inside a worktable in the Tempdb database . While an Eager Spool is used to prevent the Halloween problem, the Lazy Spool is instead used to store data that will later be needed again when running a query . So, when a Lazy spool is used?  ...
Read more

SQL Server, datetime vs. datetime2

Image
Hi guys, This time, in this post we will speak about datetime and datetime2 datatypes. We will talk about the pros and cons. We will talk about all the differences between the two types of data. Write me in the comments which type of data do you prefer between the two. Enjoy the reading!     Datetime Vs. Datetime2 First of all, it must be said that the datatype datetime is certainly the best known native format for storing a date in SQL Server. However, Microsoft has also been introducing a new type of data for many years now, also used to manage dates. It is in fact from SQL server 2008 that it is present in datetime2 data type. This "new" type of data was born primarily to have higher accuracy . Let's in fact talk about ... precision ! Datetime Certainly the datetime data type is not very precise and certainly there are many cases in which its precision is simply not sufficient. Let's see it with an example. Open the SQL Management studio (SSMS) and write the fo...
Read more

La clausola NOLOCK. Approfondiamo e facciamo chiarezza!

Image
Ehi ragazzi! La domanda è di quelle importanti! Siete pronti per scoprire tutto ma proprio tutto sulla clausola NOLOCK?   Se la riposta è si, allora seguitemi! Oggi ne spiegheremo il funzionamento in maniera approfondita. Faremo una introduzione al concetto di tipi di LOCK Ed infine vedremo come mai una SELECT ne può bloccare un'altra. Non vi nascondo che il motivo di questo articolo è fare chiarezza. Su questo argomento si sentono spesso affermazioni “curiose” ma tra le tante questa è la prima: “ uso la SELECT con l’opzione WITH(LOCK) così non blocco le altre tabelle ”.   Ma, è proprio vero quanto affermato? Beh oggi facciamo il punto!   Buona lettura!     La clausola NOLOCK Diciamo subito per iniziare che cos’è la NOLOCK. NOLOCK è il nome di una clausola che è possibile specificare all’interno di uno statement di tipo SELECT utilizzando la seguente sintassi: SELECT * FROM TABELLA WITH (NOLOCK) Si noti che non possiamo ...
Read more