A deadlock occurs when neither of two transactions can be committed because they each have a lock on a resource needed by the other.
When several transactions are awaiting one another's lock releases, a deadlock results. A detection mechanism normally handles deadlocks automatically. Timeouts can be used by the system to identify transactions that have been held up for too long and may be in a deadlock. Alternately, it explicitly keeps track of a waits-for graph and routinely scans for cycles. By cancelling one of the transactions implicated in the deadlock, the system ends the impasse.
Performance lags brought on by lock conflicts are the biggest issue with programme design caused by locking. Transaction throughput drops if too many transactions request competing locks. Lock thrashing is the term for this. The number of ongoing transactions must be reduced by aborting them in order to solve it in a running system.
To learn more about deadlock here,
https://brainly.com/question/16901684
#SPJ4
