Transactions and Concurrency

Transactions and Concurrency Susan B. Davidson University of Pennsylvania CSE330 – Database Management Systems Some slide content derived from Ramakrishnan & Gehrke

From Queries to Updates • We’ve spent a lot of time talking about querying data • Yet updates are a really major part of many DBMS applications • Standard notion for updates is a transaction: • Sequence of read and write operations on data items that logically functions as one unit of work • If it succeeds, the effects of all write operations persist (commit); if it fails, no effects persist (abort) • These guarantees are made despite concurrent activity in the system, and despite failures that may occur

Outline • Transactions and ACID properties: the dangers in concurrent executions (Ch. 16:1-3) • Transactions and SQL: isolation levels (Ch. 16.6) • How the database implements isolation levels (Ch. 16:4-5) • Theory of serializability

ACID Properties • Atomicity: either all of the actions of a transaction are executed, or none are. • Consistency: each transaction executed in isolation keeps the database in a consistent state (this is the responsibility of the user) • Isolation: can understand what’s going on by considering each transaction independently. Transactions are isolated from the effects of other, concurrently executing, transactions. • Durability: updates stay in the DBMS!!!

How Things Can Go Wrong • Suppose we have a table of bank accounts which contains the balance of the account • An ATM deposit of $50 to account # 1234 would be written as: • This reads and writes the account’s balance • What if two accountholders make deposits simultaneously from two ATMs? update Accounts set balance = balance + $50 where account#= ‘1234’;

Concurrent Deposits This SQL update code is represented as a sequence of read and write operations on “data items” (which for now should be thought of as individual accounts): where X is the data item representing the account with account# 1234. Deposit 1 Deposit 2 read(X.bal) read(X.bal) X.bal := X.bal + $50 X.bal:= X.bal + $10 write(X.bal) write(X.bal)

BAD! time A “Bad” Concurrent Execution Only one “action” (e.g. a read or a write) can actually happen at a time, and we can interleave deposit operations in many ways: Deposit 1 Deposit 2 read(X.bal) read(X.bal) X.bal := X.bal + $50 X.bal:= X.bal + $10 write(X.bal) write(X.bal)

GOOD! time A “Good” Execution • Previous execution would have been fine if the accounts were different (i.e. one were X and one were Y), i.e., transactions were independent • The following execution is a serial execution, and executes one transaction after the other: Deposit 1 Deposit 2 read(X.bal) X.bal := X.bal + $50 write(X.bal) read(X.bal) X.bal:= X.bal + $10 write(X.bal)

Good Executions An execution is “good” if it is serial (transactions are executed atomically and consecutively) or serializable (i.e. equivalent to some serial execution) Equivalent to executing Deposit 1 then 3, or vice versa • Why is this good? Deposit 1 Deposit 3 read(X.bal) read(Y.bal) X.bal := X.bal + $50 Y.bal:= Y.bal + $10 write(X.bal) write(Y.bal)

Transfer read(X.bal) read(Y.bal) X.bal= X.bal-$100 Y.bal= Y.bal+$100 CRASH Atomicity Problems can also occur if a crash occurs in the middle of executing a transaction: Need to guarantee that the write to X does not persist (ABORT) • Default assumption if a transaction doesn’t commit

Transactions in SQL • A transaction begins when any SQL statement that queries the db begins. • To end a transaction, the user issues a COMMIT or ROLLBACK statement. Transfer UPDATE Accounts SET balance = balance - $100 WHERE account#= ‘1234’; UPDATE Accounts SET balance = balance + $100 WHERE account#= ‘5678’; COMMIT;

Read-Only Transactions • When a transaction only reads information, we have more freedom to let the transaction execute in parallel with other transactions. • We signal this to the system by stating: SET TRANSACTION READ ONLY; SELECT * FROM Accounts WHERE account#=‘1234’; ...

Read-Write Transactions • If we state “read-only”, then the transaction cannot perform any updates. • Instead, we must specify that the transaction may update (the default): SET TRANSACTION READ ONLY; UPDATE Accounts SET balance = balance - $100 WHERE account#= ‘1234’; ... ILLEGAL! SET TRANSACTION READ WRITE; update Accounts set balance = balance - $100 where account#= ‘1234’; ...

Dirty Reads • Dirty data is data written by an uncommitted transaction; a dirty read is a read of dirty data • Sometimes we can tolerate dirty reads; other times we cannot.

“Bad” Dirty Read EXEC SQLselect balance into :bal from Accounts where account#=‘1234’; if (bal > 100) { EXEC SQLupdate Accounts set balance = balance - $100 where account#= ‘1234’; EXEC SQLupdate Accounts set balance = balance + $100 where account#= ‘5678’; } EXEC SQLCOMMIT; If the initial read (italics) were dirty, the balance could become negative!

Acceptable Dirty Read If we are just checking availability of an airline seat, a dirty read might be fine! (Why is that?) Reservation transaction: EXEC SQL select occupied into :occ from Flights where Num= ‘123’ and date=11-03-99 and seat=‘23f’; if (!occ) {EXEC SQL update Flights set occupied=true where Num= ‘123’ and date=11-03-99 and seat=‘23f’;} else {notify user that seat is unavailable}

Other Undesirable Phenomena • Unrepeatable read: a transaction reads the same data item twice and gets different values • Phantom problem: a transaction retrieves a collection of tuples twice and sees different results

Phantom Problem Example • T1: “find the students with best grades who Take either cis550-f03 or cis570-f02” • T2: “insert new entries for student #1234 in the Takes relation, with grade A for cis570-f02 and cis550-f03” • Suppose that T1 consults all students in the Takes relation and finds the best grades for cis550-f03 • Then T2 executes, inserting the new student at the end of the relation, perhaps on a page not seen by T1 • T1 then completes, finding the students with best grades for cis570-f02 and now seeing student #1234

Isolation • The problems we’ve seen are all related to isolation • General rules of thumb w.r.t. isolation: • Fully serializable isolation is more expensive than “no isolation” • We can’t do as many things concurrently (or we have to undo them frequently) • For performance, we generally want to specify the most relaxed isolation levelthat’s acceptable • Note that we’re “slightly” violating a correctness constraint to get performance!

Specifying Acceptable Isolation Levels • The default isolation level is SERIALIZABLE (as for the transfer example). • To signal to the system that a dirty read is acceptable, • In addition, there are SET TRANSACTION READ WRITE ISOLATION LEVEL READ UNCOMMITTED; SET TRANSACTION ISOLATION LEVEL READ COMMITTED; SET TRANSACTION ISOLATION LEVEL REPEATABLE READ;

READ COMMITTED • Forbids the reading of dirty (uncommitted) data, but allows a transaction T to issue the same query several times and get different answers • No value written by T can be modified until T completes • For example, the Reservation example could also be READ COMMITTED; the transaction could repeatably poll to see if the seat was available, hoping for a cancellation

REPEATABLE READ • What it is NOT: a guarantee that the same query will get the same answer! • However, if a tuple is retrieved once it will be retrieved again if the query is repeated • For example, suppose Reservation were modified to retrieve all available seats • If a tuple were retrieved once, it would be retrieved again (but additional seats may also become available)

Summary of Isolation Levels Level Dirty Read Unrepeatable Read Phantoms READ UN- Maybe Maybe Maybe COMMITTED READ No Maybe Maybe COMMITTED REPEATABLE No No Maybe READ SERIALIZABLE No No No

Implementing Isolation Levels • One approach – use locking at some level (tuple, page, table, etc.): • each data item is either locked (in some mode, e.g. shared or exclusive) or is available (no lock) • an action on a data item can be executed if the transaction holds an appropriate lock • consider granularity of locks – how big of an item to lock • Larger granularity = fewer locking operations but more contention! • Appropriate locks: • Before a read, a shared lock must be acquired • Before a write, an exclusive lock must be acquired

Mode of Data Item None Shared Exclusive Shared Y Y N Exclusive Y N N Request mode{ Lock Compatibility Matrix Locks on a data item are granted based on a lock compatibility matrix: When a transaction requests a lock, it must wait (block) until the lock is granted

Locks Prevent “Bad” Execution If the system used locking, the first “bad” execution could have been avoided: Deposit 1 Deposit 2 xlock(X) read(X.bal) {xlock(X) is not granted} X.bal := X.bal + $50 write(X.bal) release(X) xlock(X) read(X.bal) X.bal:= X.bal + $10 write(X.bal) release(X)

Locks are not enough… Deposit 1 Deposit 2 slock(X) read(X.bal) slock(X) read(X.bal) release(X) release(X) X.bal:= X.bal + $10 X.bal := X.bal + $50 xlock(X) write(X.bal) release(X) xlock(X) write(X.bal) release(X)

Lock Types and Read/Write Modes When we specify “read-only”, the system only uses shared-mode locks • Any transaction that attempts to update will be illegal When we specify “read-write”, the system may also acquire locks in exclusive mode • Obviously, we can still query in this mode

Isolation Levels and Locking READ UNCOMMITTED allows queries in the transaction to read data without acquiring any lock For updates, exclusive locks must be obtained and held to end of transaction READ COMMITTED requires a read-lock to be obtained for all tuples touched by queries, but it releases the locks immediately after the read Exclusive locks must be obtained for updates and held to end of transaction

Isolation levels and locking, cont. REPEATABLE READplaces shared locks on tuples retrieved by queries, holds them until the end of the transaction Exclusive locks must be obtained for updates and held to end of transaction SERIALIZABLE places shared locks on tuples retrieved by queries as well as the index, holds them until the end of the transaction Exclusive locks must be obtained for updates and held to end of transaction Holding locks to the end of a transaction is called “strict” locking

Theory of Serializability • A schedule of a set of transactions is a linear ordering of their actions • e.g. for the simultaneous deposits example: R1(X.bal) R2(X.bal) W1(X.bal) W2(X.bal) • A serial schedule is one in which all the steps of each transaction occur consecutively • A serializable schedule is one which is equivalent to some serial schedule (i.e. given any initial state, the final state is the same as one produced by some serial schedule) • The example above is neither serial nor serializable

Questions to Address • Given a schedule S, is it serializable? • How can we "restrict" transactions in progress to guarantee that only serializable schedules are produced?

When Actions Conflict • Consider a schedule S in which there are two consecutive actions Ii and Ij of transactions Ti and Tj respectively • If Ii and Ij refer to different data items, then swapping (i.e. reordering) Ii and Ij does not matter • If Ii and Ij refer to the same data item Q, then swapping Ii and Ij matters if and only if one of the actions is a write • Ri(Q) Wj(Q) produces a different final value for Q than Wj(Q) Ri(Q)

Testing for Serializability • Given a schedule S, we can construct a directed graph G=(V,E) called a precedence graph • V : all transactions in S • E : Ti Tj whenever an action of Ti precedes and conflicts with an action of Tj in S • Theorem: A schedule S is conflict serializable if and only if its precedence graph contains no cycles • Note that testing for a cycle in a digraph can be done in time O(|V|2)

T1 T2 T3 R(X,Y,Z) R(X) W(X) R(Y) W(Y) R(Y) R(X) W(Z) T1 T2 T3 Cyclic: Not serializable. An Example

Locking and Serializability • We said that a transaction must hold all locks until it terminates (a condition called strict locking) • It turns out that this is crucial to guarantee serializability • Note that the first (bad) example could have been produced if transactions acquired and immediately released locks.

Well-Formed, Two-Phased Transactions • A transaction is well-formed if it acquires at least a shared lock on Q before reading Q or an exclusive lock on Q before writing Q and doesn’t release the lock until the action is performed • Locks are also released by the end of the transaction • A transaction is two-phased if it never acquires a lock after unlocking one • i.e., there are two phases: a growing phase in which the transaction acquires locks, and a shrinking phase in which locks are released

Two-Phased Locking Theorem • If all transactions are well-formed and two-phase, then any schedule in which conflicting locks are never granted ensures serializability • i.e., there is a very simple scheduler! • However, if some transaction is not well-formed or two-phase, then there is some schedule in which conflicting locks are never granted but which fails to be serializable • i.e., one bad apple spoils the bunch

Summary • Transactions are all-or-nothing units of work guaranteed despite concurrency or failures in the system. • Theoretically, the “correct” execution of transactions is serializable (i.e. equivalent to some serial execution). • Practically, this may adversely affect throughput  isolation levels. • With isolation levels, users can specify the level of “incorrectness” they are willing to tolerate.

Transactions and Concurrency