M.C.A. Semester - II
620006 – Databse Management System - II
Assignment – 3 (Database Backup and Recovery)
------------------------------------------------------------------------------------------------------------
Q-1. Explain the Following Questions in detail.
1). What is Database Backup? Why backup is needed? Explain Types of Database Backup?
2). Discuss different types of database failures that may occur in database environment?
3). What is Recovery? Explain types of Recovery? (Forward and Backward Recovery)
4). What is Log Based Recovery? Explain all Log based Recovery technique with Example?
5). What is Checkpoint? How is checkpoint information is used in recovery operation following System crash explain with example?
6). Describe Shadow Paging Recovery Technique? Under what circumstances does it
not require a transaction log? List advantages and disadvantage of shadow paging?
not require a transaction log? List advantages and disadvantage of shadow paging?
7). What is Database Buffer? Explain Buffer Management technique used in database recovery?
8). Table : Transaction T1
Time shot
|
Transaction Step
|
Actions
|
Time-1
|
Read (A,a1)
|
Read current employee’s load balance.
|
Time-2
|
a1:= a1 – 500
|
Debit the Account by INR 500
|
Time-3
|
Write(A,a1)
|
Write the new loan balance
|
Time-4
|
Read (B,b1)
|
Read the current account payable balance
|
Time-5
|
b1 := b1 + 500
|
Credit the account balance by INR 500
|
Time-6
|
Write(B,b1)
|
Write the new balance
|
Time-7
|
Commit T1
|
Commit entire transaction
|
Consider
the transaction as above Table to create log entries for immediate
updates. Suppose, that a failure occurs just after the transaction log
record for action Write(B,b1) has been Written.
a). Show the Contents of transaction log at the time of failure.
b). What action is necessary and Why?
c). What are the Resulting value of A and B?
9).
Consider the transaction as above Table to create log entries for
deferred updates. Suppose, that a failure occurs just after the record
is written to the transaction log?
a). Show the Contents of transaction log at the time of failure.
b). What action is necessary and Why?
c). What are the Resulting value of A and B?
10). Consider the following Transaction Step for Immediate Update Technique.
Initial Value of A=1000, B=2000 and C=700.
Initial Value of A=1000, B=2000 and C=700.
Transaction Step
|
Actions
|
Begin Transaction T1
|
Start Transaction T1
|
Read (A)
|
Read Value A
|
A:= A – 50
|
Deduct 50 from A
|
Write(A)
|
Write the new value of A
|
Read (B)
|
Read the Value of B
|
B := B + 50
|
Add 50 to B
|
Write(B)
|
Write the new value of B
|
End Transaction
|
Commit entire transaction T1
|
Begin Transaction T2
|
Start Transaction T2
|
Read (C)
|
Read Value of C
|
C:=C-100
|
Deduct 100 from C
|
Write (C)
|
Write the new value for C
|
End Transaction T2
|
Commit Entrire transaction T2
|
- Create Log entries for Above transaction steps.
- Let us Assume that the crash occurs just after the log record for step write(B) of transaction T1 then
- What action is necessary and why?
- What are the resulting value of A and B
- Let us Assume that Crash comes just after the log record for the Wrtite(C) of Transaction T2 then
- Finally, Let us Assume that Crash occurs just after the log record for Transaction T2,Commit
- What action is necessary and Why?
- What are the resulting value of A and B?
11). Consider the following log entries.
Log Entries
|
T1,Start
|
T1,A,5000,4500
|
T1,Commit
|
Checkpoint
|
T2,Start
|
T2,B,1000,1050
|
T2,C,4000,3900
|
T2,Commit
|
T3,Start
|
T3,D,2500,2000
|
T4,Start
|
T4,E,1000,1200
|
T4,F,1200,1400
|
Suppose, that system crash is occurs after the last write entry of F of transaction T4
then What action is necessary and why?
then What action is necessary and why?
----------------------------------------------------------------------------------------------------------------------------------------------
Q-2. Differentiate the following :
1). Deferred Update Vs Immediate Update
2). Rollforward Vs Rollback
3). Stable Storage Vs Disk
4). Volatile Storage Vs Non-Volatile Storage
5). System Crash Vs Media Failure
-----------------------------------------------------------------------------------------------------------------------------------------------
Q-3. Answer the following in One line?
1) If the buffer pool is large enough that uncommited data are never forced to disk, is UNDO still necessary?
2) How about REDO?
3) If updates are always forced to disk when a transaction is performed, is UNDO still necessary?
4) How about REDO?
5) After a soft crash, where in the log should Analysis start?
------------------------------------------------------------------------------------------------------------
0 comments:
Post a Comment