Transaction and concurrency control
Download as PPTX, PDF0 likes1,020 views
The document discusses transaction processing and concurrency control techniques in databases. It defines transactions and their ACID properties. It describes different states of transactions, nested transactions, locking techniques including shared and exclusive locks, and their compatibility. It also discusses optimistic concurrency control and timestamp ordering for concurrency control and avoiding deadlocks.
Transaction and concurrency control
- 1. - A n i l S h r e s t h a ( 0 7 0 / B C T / 0 1 ) - B i j a y S a h a n i ( 0 7 0 / B C T / 0 5 ) - B i m a l S h r e s t h a ( 0 7 0 / B C T / 1 0 ) - D e s h b h a k t a K h a n a l ( 0 7 0 / B C T / 1 3 ) 17 March 2017 1 TRANSACTION & CONCURRENCY CONTROL
- 2. What is a Transaction? 17 March 2017 2 Logical unit of work in database An entire program A portion of program A single command Must be either entirely completed or aborted No intermediate steps are acceptable Successful transactions change database from one consistent state to another
- 3. ACID Properties Of Transaction 17 March 2017 3 Atomic – All or nothing All parts of transaction must be completed and committed or must be aborted and rolled back Consistent Conducts transactions in serial order Important in multi-user and distributed databases
- 4. ACID Properties Of Transaction 17 March 2017 4 Isolation There may be execution of multiple transaction and each transaction must be unware of each other Durability Changes made to database persists even if there are system failures
- 5. States Of Transaction 17 March 2017 5 Active Partially Committed Committed Failed Aborted
- 6. Nested Transaction 17 March 2017 6 A transaction that is created inside another transaction While the nested (child) transaction is active, the parent transaction may not perform any operations other than to commit or abort, or to create more child transactions When a parent aborts, all of its sub-transactions are aborted When a sub-transaction aborts, parent can decide whether to abort or not
- 7. Lock 17 March 2017 7 A lock guarantees exclusive use of a data item to a current transaction To access data item (lock acquire) After completion of transaction(release lock) All data item must be accessed in mutual exclusive manner
- 8. Types of Lock 17 March 2017 8 Shared lock Exclusive lock
- 9. Shared Lock 17 March 2017 9 When we want read the data item value at that time shared lock is used Lock-s
- 10. Exclusive Lock 17 March 2017 10 It is used for both read and write Lock-x
- 11. Compatibility Between Lock Modes 17 March 2017 11 shared exclusive shared true false exclusive false false
- 12. Example of lock 17 March 2017 12 T1 T2 Lock-x(B) R(B) B-50 W(B) Unlock(B) lock-s(B) R(B) unlock(B)
- 13. Optimistic Concurrency Control 17 March 2017 13 Optimistic concurrency control does not involve locking rows when reading Optimistic concurrency control (OCC) helps increase database performance Fewer resources are used to hold the locks during the update process Records are locked for a shorter length of time
- 14. Method For Concurrent Control 17 March 2017 14 Deadlock: A state in which each member of a group of transactions is waiting for some other member to release a lock. Prevention: Lock all the objects used by a transaction when it starts not a good way.
- 15. Method For Concurrent Control 17 March 2017 15 Drawbacks of locking: Lock maintenance represents an overhead that is not present in systems that do not support concurrent access to shared data. Deadlock: Deadlock prevention reduces concurrency
- 16. Method For Concurrent Control 17 March 2017 16 Timestamp ordering: Each transaction is assigned a unique timestamp values when it starts Timestamp defines its position in the time sequence of transaction
- 17. Method For Concurrent Control 17 March 2017 17 Timestamp ordering write rule: if (Tc ≥ maximum read timestamp on D && Tc > write timestamp on committed version of D) perform write operation on tentative version of D with write timestamp Tc else /*write is too late*/ abort transaction Tc
- 18. Method For Concurrent Control 17 March 2017 18 Timestamp ordering read rule: If (Tc> write timestamp on committed version of D) { let Dselected be the version of D with the maximum write timestamp ≤ Tc if (Dselected is committed) perform read operation on the version Dselected else wait until the transaction that made version Dselected commits or aborts then reapply the read rule } Else abort transaction Tc
- 19. Method For Concurrent Control 17 March 2017 19 Multi-version timestamp ordering: A list of old committed versions as well as tentative versions is kept for each object. Read operations that arrive too late need not be rejected.