Query Optimization

1. Query Optimization Chapter 15

2. Query Evaluation Query Query Parser Parsed query Query Optmizer Plan Generator Plan Cost Estimator Catalog Manager Evaluation Plan Query Plan Evaluator

3. Query Optimization • It is responsible for identifying an efficient execution plan. • A query is essentially treated as a , ,algebra expression, with the remaining operations. • Optimizing such a relational algebra expression involves two basic steps: • Enumerating alternative plans (Typically a subset of all possible plans). • Needs equivalence rules • Estimating the cost of each enumerated plan and choosing the plan with the lowest estimated cost.

4. Query Evaluation Plan • A Query Evaluation Plan (or simply plan): Tree of extended R.A. ops, with choice of algorithm for each operation. • Two main issues: • For a given query, what plans are considered? • Algorithm to search plan space for cheapest (estimated) plan. • How is the cost of a plan estimated? • Ideally: Want to find best plan. Practically: Avoid worst plans!

5. sname rating > 5 bid=100 sid=sid Sailors Reserves (On-the-fly) sname (On-the-fly) rating > 5 bid=100 (Simple Nested Loops) sid=sid Sailors Reserves Query Evaluation Plans SELECT S.sname FROM Reserves R, Sailors S WHERE R.sid=S.sid AND R.bid=100 AND S.rating>5 • Consider the following query: • The same query in R.A.: sname(bid =100  rating > 5 (Reserves Sailors)) R.A. tree:Plan: file scan file scan

6. C A B Pipelined vs Materialized Evaluation • When a query is composed of several operators, the result is sometimes pipelined to another operator without creating a temporary file. • It is a control strategy governing the rate at which different joins in the plan proceed. • Has lower ovehead cost. • If the output of an operator is saved in a temporary file for processing by the next operator we say that tuples are materialized. • Example: Results of 1st join pipelined into 2nd

7. (On-the-fly) sname (On-the-fly) rating > 5 bid=100 (Simple Nested Loops) sid=sid Sailors Reserves Motivating Example SELECT S.sname FROM Reserves R, Sailors S WHERE R.sid=S.sid AND R.bid=100 AND S.rating>5 • Cost: 500+500*1000 I/Os • By no means the worst plan! • Misses several opportunities: selections could have been `pushed’ earlier, no use is made of any available indexes, etc. • Goal of optimization: To find more efficient plans that compute the same answer. Plan:

8. (On-the-fly) sname (Sort-Merge Join) sid=sid (Scan; (Scan; write to write to rating > 5 bid=100 temp T2) temp T1) Reserves Sailors Alternative Plans 1 (No Indexes) • Main difference: push selects. • With 5 buffers, cost of plan: • Scan Reserves (1000) + write temp T1 (10 pages, if we have 100 boats, uniform distribution). • Scan Sailors (500) + write temp T2 (250 pages, if we have 10 ratings). • Sort T1 (2*2*10), sort T2 (2*3*250), merge (10+250) • Total: 3560 page I/Os. • If we used BNL join, join cost = 10+4*250, total cost = 2770. • If we `push’ projections, T1 has only sid, T2 only sid and sname: • T1 fits in 3 pages, cost of BNL drops to under 250 pages, total < 2000.

9. (On-the-fly) sname Alternative Plans 2With Indexes (On-the-fly) rating > 5 • With clustered index on bid of Reserves, we get 100,000/100 = 1000 tuples on 1000/100 = 10 pages. • INL with pipelining (outer is not materialized). (Index Nested Loops, with pipelining ) sid=sid (Use hash Sailors bid=100 index; do not write result to temp) Reserves • Projecting out unnecessary fields from outer doesn’t help. • Join column sid is a key for Sailors. • At most one matching tuple, unclustered index on sid OK. • Decision not to push rating>5 before the join is based on • availability of sid index on Sailors. • Cost: Selection of Reserves tuples (10 I/Os); for each, • must get matching Sailors tuple (1000*1.2); total 1210 I/Os.

10. Alternative Plans Considered • Relational Algebra equivalences play a central role in identifying alternative plans. E.g: • Selections and cross products combined into joins. • Joins can be reordered. • Selections and projections can be “pushed”. • In particular, join ordering is important. • As the number of joins increases, the number of alternative plans increases rapidly. • Thus, it’s necessary to prune the search space. • Usually search is restricted to left-deep plans only,since they allow fully pipelined plans.

11. Left-Deep Plans • Consider a query of the form A B C D D D C C A B C D A B A B linear tree bushy tree linear tree, also left-deep tree

12. Equivalence Rules 1. Conjunctive selection operations can be deconstructed into a sequence of individual selections. 2. Selection operations are commutative. 3. Only the last in a sequence of projection operations is needed, the others can be omitted. • Selections can be combined with Cartesian products and theta joins. • (E1X E2) = E1 E2 • 1(E12 E2) = E11 2E2

13. Pictorial Depiction of Equivalence Rules

14. Equivalence Rules (Cont.) 5. Theta-join operations (and natural joins) are commutative.E1  E2 = E2 E1 6. Natural join operations are associative: (E1 E2) E3 = E1 (E2 E3)

15. Equivalence Rules (Cont.) • The selection operation distributes over the theta join operation under the following two conditions: • When all the attributes in 0 involve only the attributes of one of the expressions (E1) being joined.0E1  E2) = (0(E1))  E2 • When 1 involves only the attributes of E1 and2 involves only the attributes of E2. 1 E1 E2) = (1(E1))  ( (E2))

16. Equivalence Rules (Cont.) 8. The projections operation distributes over the theta join operation as follows: (a) if  involves only attributes from L1 L2: (b) Consider a join E1  E2. • Let L1 and L2 be sets of attributes from E1 and E2, respectively. • Let L3 be attributes of E1 that are involved in join condition , but are not in L1 L2, and • let L4 be attributes of E2 that are involved in join condition , but are not in L1 L2.

17. Equivalence Rules (Cont.) • The set operations union and intersection are commutative E1 E2 = E2 E1E1 E2 = E2 E1 • (set difference is not commutative). • Set union and intersection are associative. (E1 E2)  E3 = E1 (E2  E3)(E1 E2)  E3 = E1 (E2  E3) • The selection operation distributes over ,  and –. (E1 – E2) = (E1) – (E2)and similarly for  and  in place of –Also: (E1 – E2) = (E1) – E2and similarly for in place of –, but not for  12. The projection operation distributes over union L(E1 E2) = (L(E1))  (L(E2))

18. Transformation Example • Query: Find the names of sailors who reserved a red boat after 19/3/2004.sname(color = “red”  day > 19/3/2004(boats (reserves sailors))) • Transformation using join associativity and comm.: sname(( color = “red”  day > 19/3/2004 (reserves boats) sailors) • Apply the “perform selections early” rule, resulting in the subexpression  color = “red”(boats)  day > 19/3/2004 (reserves ) • Thus a sequence of transformations can be useful

19. Translating SQL Queries into R.A. • As an example consider the following SQL query: SELECT sname, age FROM Sailors WHERE rating > SELECT max(rating) FROM Sailors WHERE age = 30 • Decompose into 2 blocks: SELECT max(rating) FROM Sailors WHERE age = 30 SELECT sname, age FROM Sailors WHERE rating > constant sname,age (rating>constant (Sailors)) Max(rating (age = 30 (Sailors))

20. Query Blocks: Units of Optimization SELECT S.sname FROM Sailors S WHERE S.age IN (SELECT MAX (S2.age) FROM Sailors S2 GROUP BY S2.rating) • An SQL query is parsed into a collection of queryblocks, and these are optimized one block at a time. • Nested blocks are usually treated as calls to a subroutine, made once per outer tuple. (This is an over-simplification, but serves for now.) Outer block Nested block • For each block, the plans considered are: • All available access methods, for each reln in FROM clause. • All left-deep join trees(i.e., all ways to join the relations one-at-a-time, with the inner reln in the FROM clause, considering all reln permutations and join methods.)

21. Estimating the cost of a plan • For each plan considered, must estimate cost: • Must estimate costof each operation in plan tree. • Depends on input cardinalities. • We’ve already discussed how to estimate the cost of operations (sequential scan, index scan, joins, etc.) • Must estimate size of result for each operation in tree! • Use information about the input relations. • For selections and joins, assume independence of predicates. • We’ll discuss the System R cost estimation approach. • Very inexact, but works ok in practice. • More sophisticated techniques known now.

22. Highlights of System R Optimizer • Impact: • Most widely used currently; works well for < 10 joins. • Cost estimation: Approximate art at best. • Statistics, maintained in system catalogs, used to estimate cost of operations and result sizes. • Considers combination of CPU and I/O costs. • Plan Space: Too large, must be pruned. • Only the space of left-deep plans is considered. • Left-deep plans allow output of each operator to be pipelinedinto the next operator without storing it in a temporary relation. • Cartesian products avoided.

23. Estimating result sizes • Consider a query block: • Every term in WHERE eliminates some of the result tuples. • A reduction factor is associated with each term. • The number of tuples in the result is estimated as the maximum size (i.e. product of cardinalities) times the product of reduction factors. SELECT attribute list FROM relation list WHERE term1 term2 ...  termn

24. High(I) - value High(I) – Low(I) Computing reduction factors • column = value: • if there is an index I on column : 1/Nkeys(I); • if no index and no statistics about the distinct values: 1/10 (arbitrarily). • column > value: • If there is an index: • No index, not of arithmetic type: 1/2 (arbitrarily).

25. 1 max(Nkeys(I1), NKeys(I2)) 1 Nkeys(I) Reduction factors • column1 = column2: • if there are indexes I1 and I2 on column1 and column2 respectively: • if only one of the columns have an index: • if no index: 1/10 arbitrarily

26. Estimation of the Size of Joins • The Cartesian product r x s contains nr .nstuples; each tuple occupies sr + ssbytes. • If R  S = , then rs is the same as r x s. • If R  S is a key for R, then a tuple of s will join with at most one tuple from r • therefore, the number of tuples in r s is no greater than the number of tuples in s. • If R  Sin S is a foreign key in S referencing R, then the number of tuples in rs is exactly the same as the number of tuples in s. • The case for R  S being a foreign key referencing S is symmetric. • In the example query reserves sailors, sid in reserves is a foreign key of sailors • hence, the result has exactly nreserves tuples.

27. Estimation of the Size of Joins (Cont.) • If R  S = {A} is not a key for R or S.If we assume that every tuple t in R produces tuples in R S, the number of tuples in RS is estimated to be:If the reverse is true, the estimate obtained will be:The lower of these two estimates is probably the more accurate one.

28. Enumeration of Alternative Plans • There are two main cases: • Single-relation plans • Multiple-relation plans • For queries over a single relation, queries consist of a combination of selects, projects, and aggregate ops: • Each available access path (file scan / index) is considered, and the one with the least estimated cost is chosen. • The different operations are essentially carried out together (e.g., if an index is used for a selection, projection is done for each retrieved tuple, and the resulting tuples are pipelined into the aggregate computation).

29. Cost Estimates for Single-Relation Plans • Index I on primary key matches selection: • Cost is Height(I)+1 for a B+ tree, about 1.2 for hash index. • Clustered index I matching one or more selects: • (NPages(I)+NPages(R)) * product of RF’s of matching selects. • Non-clustered index I matching one or more selects: • (NPages(I)+NTuples(R)) * product of RF’s of matching selects. • Sequential scan of file: • NPages(R). • Note:Typically, no duplicate elimination on projections! (Exception: Done on answers if user says DISTINCT.)

30. Example SELECT S.sid FROM Sailors S WHERE S.rating=8 • If we have an index on rating: • (1/NKeys(I)) * NTuples(R) = (1/10) * 40000 tuples retrieved. • Clustered index: (1/NKeys(I)) * (NPages(I)+NPages(R)) = (1/10) * (50+500) pages are retrieved. (This is the cost.) • Unclustered index: (1/NKeys(I)) * (NPages(I)+NTuples(R)) = (1/10) * (50+40000) pages are retrieved. • If we have an index on sid: • Would have to retrieve all tuples/pages. With a clustered index, the cost is 50+500, with unclustered index, 50+40000. • Doing a file scan: • We retrieve all file pages (500).

31. D D C C D B A C B A B A Queries Over Multiple Relations • Fundamental decision in System R: only left-deep join treesare considered. • As the number of joins increases, the number of alternative plans grows rapidly; we need to restrict the search space. • Left-deep trees allow us to generate all fully pipelined plans. • Intermediate results not written to temporary files. • Not all left-deep trees are fully pipelined (e.g., SM join).

32. Enumeration of Left-Deep Plans • Left-deep plans differ only in the order of relations, the access method for each relation, and the join method for each join. • Enumerated using N passes (if N relations joined): • Pass 1: Find best 1-relation plan for each relation. • Pass 2: Find best way to join result of each 1-relation plan (as outer) to another relation. (All 2-relation plans.) • Pass N: Find best way to join result of a (N-1)-relation plan (as outer) to the N’th relation. (All N-relation plans.) • For each subset of relations, retain only: • Cheapest plan overall, plus • Cheapest plan for each interesting order of the tuples.

33. Enumeration of Plans (Contd.) • ORDER BY, GROUP BY, aggregates etc. handled as a final step, using either an `interestingly ordered’ plan or an additonal sorting operator. • An N-1 way plan is not combined with an additional relation unless there is a join condition between them, unless all predicates in WHERE have been used up. • i.e., avoid Cartesian products if possible. • In spite of pruning plan space, this approach is still exponential in the # of tables.

34. sname sid=sid rating > 5 bid=100 Sailors Reserves Sailors: B+ tree on rating Hash on sid Reserves: B+ tree on bid Example 1 • Pass1: • Sailors: B+ tree matches rating>5, and is probably cheapest. However, if this selection is expected to retrieve a lot of tuples, and index is unclustered, file scan may be cheaper. • Still, B+ tree plan kept (because tuples are in rating order). • Reserves: B+ tree on bid matches bid=100; cheapest. • Pass 2: • We consider each plan retained from Pass 1 as the outer, and consider how to join it with the (only) other relation. • e.g., Reserves as outer: Hash index can be used to get Sailors tuples • that satisfy sid = outer tuple’s sid value.

35. Example 2 sid,count(*) as numres SELECT S.sid, COUNT(*) As numres FROM Boats B, Reserves R, Sailors S WHERE R.sid = S.sid and B.bid = R.bid and B.color = ‘red GROUP BY S.sid • Available Indexes : GROUP BYsid sid=sid Sailors bid=bid Sailors: B+ tree on sid Hash on sid Reserves: B+ tree on sid B+ tree on bid (clustered) Boats: B+ tree on color Hash on color Reserves color = red Boats

36. Pass 1: Find best plans for each file. • Sailors: File scan • Reserves: File scan • Boats: hash index on color, B+ tree index on color • Pass 2: Possible joins are considered. • Reserves Boats • Reserves Sailors • Sailors Boats • Sailors Reserves • Boats (access B+ tree) Sailors • Boats (access B+ tree) Reserves • Boats (access hash) Sailors • Boats (access hash) Reserves • For each such pair, consider every join method and for each join method consider every access path for the inner relation. • For each pair keep the cheapest of the plans & the cheapest plan that generates tuples in sorted order

37. Pass 3: For each plan retained in pass 2, taken as outer relation, consider how to join remaining relation as inner one. • Example plan for this pass: Boats (via hash) Reserves (via B+tree) (sort-merge) Result is joined with Sailors (via B+ tree) (sort merge) • GROUP BY is considered after all joins. It requires sorting on sid.

38. SELECT S.sname FROM Sailors S WHERE EXISTS (SELECT * FROM Reserves R WHERE R.bid=103 AND R.sid=S.sid) Nested Queries • Nested block is optimized independently, with the outer tuple considered as providing a selection condition. • Outer block is optimized with the cost of `calling’ nested block computation taken into account. • Implicit ordering of these blocks means that some good strategies are not considered. The non-nested version of the query is typically optimized better. Nested block to optimize: SELECT * FROM Reserves R WHERE R.bid=103 AND S.sid= outer value Equivalent non-nested query: SELECT S.sname FROM Sailors S, Reserves R WHERE S.sid=R.sid AND R.bid=103

39. Summary • Query optimization is an important task in a relational DBMS. • Must understand optimization in order to understand the performance impact of a given database design (relations, indexes) on a workload (set of queries). • Two parts to optimizing a query: • Consider a set of alternative plans. • Must prune search space; typically, left-deep plans only. • Must estimate cost of each plan that is considered. • Must estimate size of result and cost for each plan node. • Key issues: Statistics, indexes, operator implementations.

40. Summary (Contd.) • Single-relation queries: • All access paths considered, cheapest is chosen. • Issues: Selections that match index, whether index key has all needed fields and/or provides tuples in a desired order. • Multiple-relation queries: • All single-relation plans are first enumerated. • Selections/projections considered as early as possible. • Next, for each 1-relation plan, all ways of joining another relation (as inner) are considered. • Next, for each 2-relation plan that is `retained’, all ways of joining another relation (as inner) are considered, etc. • At each level, for each subset of relations, only best plan for each interesting order of tuples is `retained’.