Crowdly
Додати до Chrome
INTRO. TO DATABASE MGMT. SYST.
Шукаєте відповіді та рішення тестів для INTRO. TO DATABASE MGMT. SYST.? Перегляньте нашу велику колекцію перевірених відповідей для INTRO. TO DATABASE MGMT. SYST. в moodlenew.iitd.ac.in.
Отримайте миттєвий доступ до точних відповідей та детальних пояснень для питань вашого курсу. Наша платформа, створена спільнотою, допомагає студентам досягати успіху!
Consider the following segment of the log record containing the actions taken by transactions beforethe crash.
⟨T3 start⟩⟨T3, D, 200, 50⟩⟨T1 start⟩⟨checkpoint {T1, T3}⟩⟨T1, A, 100, 150⟩⟨T2 start⟩⟨T2, B, 200, 500⟩⟨T1, C, 50, 100⟩⟨T2 commit⟩⟨T3, D, 50, 150⟩⟨T3 abort⟩
Mark the correct statements:
In a log-based recovery scheme with Write-Ahead Logging (WAL), if the system crashes after writing a data page to disk but before writing its corresponding log record, the change is guaranteed to be redone during recovery.
In the deferred modification technique, the old value part of an update log is not required.
Which of the following is correct about WAL with force and no-force policies?
The no-steal policy is most suitable for transactions that perform a large number of updates.
A database system uses Write-Ahead Logging (WAL) to ensure atomicity and durability. Consider the following timeline for a single transaction T5:
1. ⟨T5 start⟩2. ⟨T5, X, 200, 250⟩ — log record written and flushed3. ⟨T5, Y, 300, 350⟩ — log record written and flushed4. System crashes
Which statement best describes the state of data items X and Y on disk after recovery?
Under Write-Ahead Logging, the log record for an update ⟨Ti, A, old, new⟩ must be forced to stable storage before the corresponding data page A is written to disk.
Consider a system where the log contains the following records (in order):
⟨T1 start⟩ ⟨T1, X, 10, 20⟩ ⟨T2 start⟩ ⟨checkpoint {T1, T2}⟩ ⟨T2, Y, 30, 40⟩ (crash)
Which of the following is true immediately after the crash, under recovery with WAL?
According to the "Steal/Force" policies, which combination guarantees that no recovery is needed after every transaction commit?