' & $ Chapter 14: Concurrency Control · Lock-Based Protocols · Timestamp-Based Protocols · Validation-Based Protocols · Multiple Granularity · Multiversion Schemes · Deadlock Handling · Insert and Delete Operations · Concurrency in Index Structures Database Systems Concepts 14.1 Silberschatz, Korth and Sudarshan c 1997 ' & $ Lock-Based Protocols · A lock is a mechanism to control concurrent access to a data item · Data items can be locked in two modes : 1. exclusive (X) mode. Data item can be both read as well as written. X-lock is requested using lock-X instruction. 2. shared (S) mode. Data item can only be read. S-lock is requested using lock-S instruction. · Lock requests are made to concurrency-control manager. Transaction can proceed only after request is granted. Database Systems Concepts 14.2 Silberschatz, Korth and Sudarshan c 1997 ' & $ Lock-Based Protocols (Cont.) · Lock-compatibility matrix S X S true false X false false · A transaction may be granted a lock on an item if the requested lock is compatible with locks already held on the item by other transactions · The matrix allows any number of transactions to hold shared locks on an item, but if any transaction holds an exclusive on the item no other transaction may hold any lock on the item. · If a lock cannot be granted, the requesting transaction is made to wait till all incompatible locks held by other transactions have been released. The lock is then granted. Database Systems Concepts 14.3 Silberschatz, Korth and Sudarshan c 1997 ' & $ Lock-Based Protocols (Cont.) · Example of a transaction performing locking: T2: lock-S(A); read(A); unlock(A); lock-S(B); read(B); unlock(B); display(A + B). · Locking as above is not sufficient to guarantee serializability -- if A and B get updated in-between the read of A and B, the displayed sum would be wrong. · A locking protocol is a set of rules followed by all transactions while requesting and releasing locks. Locking protocols restrict the set of possible schedules. Database Systems Concepts 14.4 Silberschatz, Korth and Sudarshan c 1997 ' & $ Pitfalls of Lock-Based Protocols · Consider the partial schedule T3 T4 lock-X(B) read(B) B := B - 50 write(B) lock-S(A) read(A) lock-S(B) lock-X(A) · Neither T3 nor T4 can make progress -- executing lock-S(B) causes T4 to wait for T3 to release its lock on B, while executing lock-X(A) causes T3 to wait for T4 to release its lock on A. · Such a situation is called a deadlock. To handle a deadlock one of T3 or T4 must be rolled back and its locks released. Database Systems Concepts 14.5 Silberschatz, Korth and Sudarshan c 1997 ' & $ Pitfalls of Lock-Based Protocols (Cont.) · The potential for deadlock exists in most locking protocols. Deadlocks are a necessary evil. · Starvation is also possible if concurrency control manager is badly designed. For example: ­ A transaction may be waiting for an X-lock on an item, while a sequence of other transactions request and are granted an S-lock on the same item. ­ The same transaction is repeatedly rolled back due to deadlocks. · Concurrency control manager can be designed to prevent starvation. Database Systems Concepts 14.6 Silberschatz, Korth and Sudarshan c 1997 ' & $ The Two-Phase Locking Protocol · This is a protocol which ensures conflict-serializable schedules. · Phase 1: Growing Phase ­ transaction may obtain locks ­ transaction may not release locks · Phase 2: Shrinking Phase ­ transaction may release locks ­ transaction may not obtain locks · The protocol assures serializability. It can be proved that the transactions can be serialized in the order of their lock points (i.e. the point where a transaction acquired its final lock). Database Systems Concepts 14.7 Silberschatz, Korth and Sudarshan c 1997 ' & $ The Two-Phase Locking Protocol (Cont.) · Two-phase locking does not ensure freedom from deadlocks · Cascading roll-back is possible under two-phase locking. To avoid this, follow a modified protocol called strict two-phase locking. Here a transaction must hold all its exclusive locks till it commits/aborts. · Rigorous two-phase locking is even stricter: here all locks are held till commit/abort. In this protocol transactions can be serialized in the order in which they commit. Database Systems Concepts 14.8 Silberschatz, Korth and Sudarshan c 1997 ' & $ The Two-Phase Locking Protocol (Cont.) · There can be conflict serializable schedules that cannot be obtained if two-phase locking is used. · However, in the absence of extra information (e.g., ordering of access to data), two-phase locking is needed for conflict serializability in the following sense: Given a transaction Ti that does not follow two-phase locking, we can find a transaction Tj that uses two-phase locking, and a schedule for Ti and Tj that is not conflict serializable. Database Systems Concepts 14.9 Silberschatz, Korth and Sudarshan c 1997 ' & $ Lock Conversions · Two-phase locking with lock conversions: ­ First Phase: can acquire a lock-S on item can acquire a lock-X on item can convert a lock-S to a lock-X (upgrade) ­ Second Phase: can release a lock-S can release a lock-X can convert a lock-X to a lock-S (downgrade) · This protocol assures serializability. But still relies on the programmer to insert the various locking instructions. Database Systems Concepts 14.10 Silberschatz, Korth and Sudarshan c 1997 ' & $ Automatic Acquisition of Locks A transaction Ti issues the standard read/write instruction, without explicit locking calls. · The operation read(D) is processed as: if Ti has a lock on D then read(D) else begin if necessary wait until no other transaction has a lock-X on D grant Ti a lock-S on D; read(D) end; Database Systems Concepts 14.11 Silberschatz, Korth and Sudarshan c 1997 ' & $ Automatic Acquisition of Locks (Cont.) · write(D) is processed as: if Ti has a lock-X on D then write(D) else begin if necessary wait until no other trans. has any lock on D, if Ti has a lock-S on D then upgrade lock on D to lock-X else grant Ti a lock-X on D write(D) end; · All locks are released after commit or abort Database Systems Concepts 14.12 Silberschatz, Korth and Sudarshan c 1997 ' & $ Graph-Based Protocols · Is an alternative to two-phase locking · Impose a partial ordering on the set D = {d1, d2, ..., dh} of all data items. ­ If di dj, then any transaction accessing both di and dj must access di before accessing dj. ­ Implies that the set D may now be viewed as a directed acyclic graph, called a database graph. · tree-protocol is a simple kind of graph protocol. Database Systems Concepts 14.13 Silberschatz, Korth and Sudarshan c 1997 ' & $ Tree Protocol E A I B C D J H F G · Only exclusive locks are allowed. · The first lock by Ti may be on any data item. Subsequently, a data item Q can be locked by Ti only if the parent of Q is currently locked by Ti . · Data items may be unlocked at any time. · A data item that has been locked and unlocked by Ti cannot subsequently be re-locked by Ti . Database Systems Concepts 14.14 Silberschatz, Korth and Sudarshan c 1997 ' & $ Graph-Based Protocols (Cont.) · The tree protocol ensures conflict serializability as well as freedom from deadlock. · Unlocking may occur earlier in the tree-locking protocol than in the two-phase locking protocol. ­ shorter waiting times, and increase in concurrency ­ protocol is deadlock-free · However,in the tree-locking protocol, a transaction may have to lock data items that it does not access. ­ increased locking overhead, and additional waiting time ­ potential decrease in concurrency · schedules not possible under two-phase locking are possible under tree protocol, and vice versa. Database Systems Concepts 14.15 Silberschatz, Korth and Sudarshan c 1997 ' & $ Timestamp-Based Protocols · Each transaction is issued a timestamp when it enters the system. If an old transaction Ti has time-stamp TS(Ti ), a new transaction Tj is assigned time-stamp TS(Tj ) such that TS(Ti ) . Each version Qk contains three data fields: ­ Content ­ the value of version Qk . ­ W-timestamp(Qk ) ­ timestamp of the transaction that created (wrote) version Qk ­ R-timestamp(Qk ) ­ largest timestamp of a transaction that successfully read version Qk · when a transaction Ti creates a new version Qk of Q, Qk 's W-timestamp and R-timestamp are initialized to TS(Ti ). · R-timestamp of Qk is updated whenever a transaction Tj reads Qk , and TS(Tj ) > R-timestamp(Qk ). Database Systems Concepts 14.33 Silberschatz, Korth and Sudarshan c 1997 ' & $ Multiversion Timestamp Ordering (Cont.) · Suppose that transaction Ti issues a read(Q) or write(Q) operation. Let Qk denote the version of Q whose write timestamp is the largest write timestamp less than or equal to TS(Ti ). 1. If transaction Ti issues a read(Q), then the value returned is the content of version Qk . 2. If transaction Ti issues a write(Q), and if TS(Ti ) < R-timestamp(Qk ), then transaction Ti is rolled back. Otherwise, if TS(Ti ) = W-timestamp(Qk ), the contents of Qk are overwritten, otherwise a new version of Q is created. · Reads always succeed; a write by Ti is rejected if some other transaction Tj that (in the serialization order defined by the timestamp values) should read Ti 's write, has already read a version created by a transaction older than Ti . Database Systems Concepts 14.34 Silberschatz, Korth and Sudarshan c 1997 ' & $ Multiversion Two-Phase Locking · Differentiates between read-only transactions and update transactions · Update transactions acquire read and write locks, and hold all locks up to the end of the transaction. That is, update transactions follow rigorous two-phase locking. ­ Each successful write results in the creation of a new version of the data item written. ­ each version of a data item has a single timestamp whose value is obtained from a counter ts counter that is incremented during commit processing. · Read-only transactions are assigned a timestamp by reading the current value of ts counter before they start execution; they follow the multiversion timestamp-ordering protocol for performing reads. Database Systems Concepts 14.35 Silberschatz, Korth and Sudarshan c 1997 ' & $ Multiversion Two-Phase Locking (Cont.) · When an update transaction wants to read a data item, it obtains a shared lock on it, and reads the latest version. When it wants to write an item, it obtains X lock on; it then creates a new version of the item and sets this version's timestamp to . · When update transaction Ti completes, commit processing occurs: ­ Ti sets timestamp on the versions it has created to ts counter + 1 ­ Ti increments ts counter by 1 · Read-only transactions that start after Ti increments ts counter will see the values updated by Ti . Read-only transactions that start before Ti increments the ts counter will see the value before the updates by Ti . Therefore only serializable schedules are produced. Database Systems Concepts 14.36 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock Handling · Consider the following two transactions: T1: write(X) T2: write(Y) write(Y) write(X) · Schedule with deadlock T1 T2 lock-X on X write(X) lock-X on Y write(Y) wait for lock-X on X wait for lock-X on Y Database Systems Concepts 14.37 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock Handling · System is deadlocked if there is a set of transactions such that every transaction in the set is waiting for another transaction in the set. · Deadlock prevention protocols ensure that the system will never enter into a deadlock state. Some prevention strategies : ­ Require that each transaction locks all its data items before it begins execution (predeclaration). ­ Impose partial ordering of all data items and require that a transaction can lock data items only in the order specified by the partial order (graph-based protocol). Database Systems Concepts 14.38 Silberschatz, Korth and Sudarshan c 1997 ' & $ More Deadlock Prevention Strategies · Following schemes use transaction timestamps for the sake of deadlock prevention alone. · wait-die scheme -- non-preemptive ­ older transaction may wait for younger one to release data item. Younger transactions never wait for older ones; they are rolled back instead. ­ a transaction may die several times before acquiring needed data item · wound-wait scheme -- preemptive ­ older transaction wounds (forces rollback) of younger transaction instead of waiting for it. Younger transactions may wait for older ones. ­ may be fewer rollbacks than wait-die scheme. Database Systems Concepts 14.39 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock prevention (Cont.) · Both in wait-die and in wound-wait schemes, a rolled back transactions is restarted with its original timestamp. Older transactions thus have precedence over newer ones, and starvation is hence avoided. · Timeout-Based Schemes : ­ a transaction waits for a lock only for a specified amount of time. After that, the wait times out and the transaction is rolled back. ­ thus deadlocks are not possible ­ simple to implement; but starvation is possible. Also difficult to determine good value of the timeout interval. Database Systems Concepts 14.40 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock Detection · Deadlocks can be described as a wait-for graph, which consists of a pair G = (V,E), ­ V is a set of vertices (all the transactions in the system) ­ E is a set of edges; each element is an ordered pair Ti Tj. · If Ti Tj is in E, then there is a directed edge from Ti to Tj , implying that Ti is waiting for Tj to release a data item. · When Ti requests a data item currently being held by Tj , then the edge Ti Tj is inserted in the wait-for graph. This edge is removed only when Tj is no longer holding a data item needed by Ti . · The system is in a deadlock state if and only if the wait-for graph has a cycle. Must invoke a deadlock-detection algorithm periodically to look for cycles. Database Systems Concepts 14.41 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock Detection (Cont.) T26 T28 T25 T27 Wait-for graph with no cycle T26 T28 T25 T27 Wait-for graph with a cycle Database Systems Concepts 14.42 Silberschatz, Korth and Sudarshan c 1997 ' & $ Deadlock Recovery · When deadlock is detected : ­ Some transaction will have to rolled back (made a victim) to break deadlock. Select that transaction as victim that will incur minimum cost. ­ Rollback ­ determine how far to roll back transaction Total rollback: Abort the transaction and then restart it. More effective to roll back transaction only as far as necessary to break deadlock. ­ Starvation happens if same transaction is always chosen as victim. Include the number of rollbacks in the cost factor to avoid starvation. Database Systems Concepts 14.43 Silberschatz, Korth and Sudarshan c 1997 ' & $ Insert and Delete Operations · If two-phase locking is used : ­ A delete operation may be performed only if the transaction deleting the tuple has an exclusive lock on the tuple to be deleted. ­ A transaction that inserts a new tuple into the database is given an X-mode lock on the tuple · Insertions and deletions can lead to the phantom phenomenon. ­ A transaction that scans a relation (eg., find all accounts in Perryridge) and a transaction that inserts a tuple in the relation (eg., insert a new account at Perryridge) may conflict in spite of not accessing any tuple in common. ­ If only tuple locks are used, non-serializable schedules can result: the scan transaction may not see the new account, yet may be serialized before the insert transaction. Database Systems Concepts 14.44 Silberschatz, Korth and Sudarshan c 1997 ' & $ Insert and Delete Operations (Cont.) · Actually, the transaction scanning the relation is reading information that indicates what tuples the relation contains, while a transaction inserting a tuple updates the same information. The information should be locked. · One solution: associate a data item with the relation, to represent the information about what tuples the relation contains. Transactions scanning the relation acquire a shared lock in the data item, while transactions inserting or deleting a tuple acquire an exclusive lock on the data item. (Note: locks on the data item do not conflict with locks on individual tuples.) · Above protocol provides very low concurrency for insertions/deletions. Index locking protocols provide higher concurrency. Database Systems Concepts 14.45 Silberschatz, Korth and Sudarshan c 1997 ' & $ Index Locking Protocol · Every relation must have at least one index. Access to a relation must be made only through one of the indices on the relation. · A transaction Ti that performs a lookup must lock all the index buckets that it accesses, in S-mode. · A transaction Ti may not insert a tuple ti into a relation r without updating all indices to r. Ti must perform a lookup on every index to find all index buckets that could have possibly contained a pointer to tuple ti, had it existed already, and obtain locks in X-mode on all these index buckets. Ti must also obtain locks in X-mode on all index buckets that it modifies. · The rules of the two-phase locking protocol must be observed. Database Systems Concepts 14.46 Silberschatz, Korth and Sudarshan c 1997 ' & $ Concurrency in Index Structures · Indices are unlike other database items in that their only job is to help in accessing data. · Index-structures are typically accessed very often, much more than other database items. · Treating index-structures like other database items leads to low concurrency. Two-phase locking on an index may result in transactions executing practically one-at-a-time. · It is acceptable to have nonserializable concurrent access to an index as long as the accuracy of the index is maintained. In particular, the exact values read in an internal node of a B+ -tree are irrelevant so long as we land up in the correct leaf node. · There are index concurrency protocols where locks on internal nodes are released early, and not in a two-phase fashion. Database Systems Concepts 14.47 Silberschatz, Korth and Sudarshan c 1997 ' & $ Concurrency in Index Structures (Cont.) · Example of index concurrency protocol: Use crabbing instead of two-phase locking on the nodes of the B+ -tree, as follows. During search/insertion/deletion: ­ First lock the root node in shared mode. ­ After locking all required children of a node in shared mode, release the lock on the node. ­ During insertion/deletion, upgrade leaf node locks to exclusive mode. ­ When splitting or coalescing requires changes to a parent, lock the parent in exclusive mode. · Above protocol can cause excessive deadlocks. Better protocols are available; see Section 14.8 for one such protocol, the B-link tree protocol. Database Systems Concepts 14.48 Silberschatz, Korth and Sudarshan c 1997