You are on page 1of 1

The three most common concurrent transaction execution problems are lost

updates,uncommitted data, and inconsistent retrievals.Concurrency control can help avoid these
problems through its ability to coordinatesimultaneous executions of transactions in a
multiprocessing database system. This isaccomplished with various methods, including the use
of a scheduler and locks.3. The DBMS component responsible for concurrency control is a
scheduler, which is aprocess that establishes the order in which the many operations during
each transactionare executed.5. Database recovery takes place through the use of data in a
given transaction log torecover a database from an inconsistent to a consistent state. A write-
ahead-log protocolensures that transaction logs are always written before any database data are
actuallyupdated, ensuring that the database can later be recovered to a consistent state.
Redundanttransaction logs ensure a physical disk failure will not impair the DBMS’s ability
torecover data. Database buffers are used to store copies of data from physical disks, whichare
themselves updated when a transaction updates data, to help speed up operations. Theupdated
copies of data in the buffers are eventually used to update the data on thephysical disks.
Checkpoints are the process in which a DBMS writes updated buffersonto a disk, with no other
requests being executed during it.Deferred-write techniques delay the update of a physical
database, only allowing thetransaction log to update. The database itself will only update after a
given transactionreaches a commit point, which uses information from the transaction log. If
thetransaction aborts, no changes will be made, preventing a need to undo the changes.Write-
through techniques differ from deferred-write techniques as they proceed to havethe database
updated immediately.

You might also like