View serializability in DBMS Theory - Step-by-Step Execution

Concept Flow - View serializability

Start with Schedule S

↓

Check if S is conflict serializable

No↓

Check if S is view serializable

No↓

Schedule S is neither view serializable nor conflict serializable

↓

End

View serializability checks if a schedule produces the same final data view as some serial schedule, even if it is not conflict serializable.

Execution Sample

DBMS Theory

T1: R(A), W(A)
T2: R(A), W(A)
Schedule S: r1(A) r2(A) w1(A) w2(A)
Check view equivalence

This schedule shows two transactions reading and writing the same data item A in an interleaved way, and we check if it is view serializable.

Analysis Table

Step	Action	Read/Write	Data Item	Result	Notes
1	T1 reads A	Read	A	Reads initial value	No prior writes
2	T2 reads A	Read	A	Reads initial value	No prior writes
3	T1 writes A	Write	A	Updates A	T1's write
4	T2 writes A	Write	A	Updates A	T2 overwrites T1's write
5	Check final value of A	View	A	Value from T2's write	Final write by T2
6	Compare with serial schedules	View	A	Equivalent to T1->T2 serial?	No, different final value
7	Compare with serial schedules	View	A	No, T2 reads initial not from T1	Read-from relation differs
8	Conclusion	View serializability	-	Schedule is NOT view serializable	No serial schedule matches both final write and read-from

💡 Schedule is NOT view serializable because although the final write on A matches T2->T1 serial schedule, T2's read sees the initial value instead of T1's write.

State Tracker

Variable	Start	After Step 3	After Step 4	Final
Value of A	Initial	Updated by T1	Updated by T2	Value from T2's write

Key Insights - 2 Insights

Why is the schedule neither conflict serializable nor view serializable?

The precedence graph has a cycle: T2→T1 from r2(A)-w1(A) conflict, T1→T2 from w1(A)-w2(A) conflict (conflict fails). No serial schedule matches read-from relations and final writes (view fails), as shown in execution_table rows 6-8.

What does it mean for two schedules to be view equivalent?

It means they read the same initial values, write the same final values, and each read operation reads the value written by the same transaction, as demonstrated in the execution_table steps 1-7.

Visual Quiz - 3 Questions

Look at the execution table, what is the value of A after step 4?

A Initial value

B Updated by T2

C Updated by T1

D Undefined

💡 Hint

Refer to execution_table row 4 where T2 writes A, overwriting T1's write.

At which step do we compare the schedule to the T1->T2 serial schedule?

A Step 7

B Step 5

C Step 3

D Step 2

💡 Hint

Check execution_table row 7 where the schedule is compared to T1->T2 serial schedule for view equivalence.

If T2's write happened before T1's write, how would the final value of A change in the variable tracker?

A Final value would be from T2's write

B Final value would be initial value

C Final value would be from T1's write

D Final value would be undefined

💡 Hint

Look at variable_tracker and consider the order of writes affecting the final value.

Concept Snapshot

View serializability means a schedule produces the same final data view as some serial schedule.
It checks read-from and final write relationships, not just conflicts.
All conflict serializable schedules are view serializable, but not vice versa.
Used to ensure correctness of concurrent transactions in databases.

Full Transcript

View serializability is a way to check if a schedule of database transactions is correct by comparing the final data seen to that of a serial schedule. The process starts by examining the schedule's read and write actions on data items. If the schedule is not conflict serializable, we check if it is view serializable by verifying if the reads and final writes match those of some serial order. For example, two transactions reading and writing the same data item A in an interleaved way: r1(A) r2(A) w1(A) w2(A) is not conflict serializable (cycle in graph) and not view serializable (read-from mismatch despite final write match). Tracking the value changes step-by-step helps confirm this. Key points include understanding that view serializability focuses on the data view (reads-from + final writes) rather than just conflicts.