1.

Critical Regions & Synchronization

1.​ Critical Section

○​ Code segment accessing shared resources (e.g., variables, files).
○​ Requires mutual exclusion to prevent race conditions.
○​ Solutions: Locks, semaphores, monitors.
2.​ Conditional Critical Regions (CCRs)
○​ Critical sections with condition variables (e.g., wait()/signal()).
○​ Process enters only if a condition is true (e.g., while (buffer_full)
wait();).
3.​ Monitors
○​ High-level synchronization construct bundling:
■​ Shared data.
■​ Procedures to access data.
■​ Condition variables for blocking.
○​ Ensures only one process executes a monitor procedure at a time.

2. Interprocess Communication (IPC)

4.​ Message Passing

○​ Processes communicate via send/receive primitives.
○​ Types:
■​ Direct (explicit sender/receiver IDs).
■​ Indirect (via mailboxes/ports).
5.​ Pipes
○​ Unidirectional communication (e.g., shell commands: cmd1 | cmd2).
○​ Named pipes (FIFOs) persist beyond process life.
6.​ Shared Memory
○​ Fastest IPC (processes map the same memory region).
○​ Requires explicit synchronization (e.g., semaphores).
7.​ Semaphores
○​ Integer variable with atomic wait() (P) and signal() (V) operations.
○​ Types:
■​ Binary (0/1 for mutual exclusion).
■​ Counting (resource pools).

3. Deadlocks
 8.​ Deadlock Characterization​
Four necessary conditions (Coffman conditions):
○​ Mutual Exclusion: Resource held exclusively.
○​ Hold and Wait: Process holds resources while waiting for others.
○​ No Preemption: Resources cannot be forcibly taken.
○​ Circular Wait: Cycle in resource allocation graph.
9.​ Deadlock Prevention​
Break one of the four conditions:
○​ Eliminate Mutual Exclusion: Allow shared access (not always possible).
○​ Remove Hold and Wait: Request all resources initially.
○​ Allow Preemption: Forcefully release resources.
○​ Avoid Circular Wait: Impose total ordering of resource requests.
10.​Deadlock Avoidance
○​ Banker’s Algorithm: Simulates resource allocation to ensure safe
states.
○​ Requires advance knowledge of max resource needs.
11.​Deadlock Detection & Recovery
○​ Detection: Use resource allocation graphs or cycle-detection
algorithms.
○​ Recovery:
■​ Process Termination: Kill deadlocked processes.
■​ Resource Preemption: Roll back and retry.
12.​Combined Approach
○​ Hybrid of prevention, avoidance, and detection (e.g., OS defaults).

4. Advanced Concepts

13.​Precedence Graphs
○​ Directed graphs to detect deadlocks in multi-threaded systems.
○​ Nodes = processes; Edges = "waits-for" relationships.
14.​Concurrent Languages
○​ Java: synchronized blocks, wait()/notify().
○​ Go: Goroutines + channels.
○​ Erlang: Message-passing between lightweight processes.
15.​Modularization
○​ Design concurrent systems as independent modules with clear
interfaces.
○​ Reduces unintended interactions.
Key Synchronization Problems

16.​Producer-Consumer
○​ Solved with bounded buffers + semaphores/monitors.
17.​Readers-Writers
○​ Prioritize readers (starvation risk) or writers (fairness).
18.​Dining Philosophers
○​ Illustrates deadlock (all pick left fork first) and solutions (resource
ordering).

Summary Table

Concept Key Mechanism Example

Critical Section Locks/Semaphores lock.acquire(); ...;

lock.release();

Monitor Encapsulated data + Java synchronized methods

procedures

Message send()/receive() MPI, Go channels

Passing

Deadlock Break Coffman conditions Request all resources upfront

Prevention

Banker’s Safe state checking Resource allocation simulation

Algorithm

1. Evolution of Operating Systems

Types of OS
 1.​ Batch Processing
○​ Jobs executed sequentially without user interaction.
○​ Example: Early mainframe systems (1950s).
2.​ Time-Sharing
○​ Multiple users interact concurrently via terminals.
○​ Example: UNIX (1970s).
3.​ Real-Time OS (RTOS)
○​ Guarantees timely response for critical tasks.
○​ Types:
■​ Hard RTOS: Strict deadlines (e.g., medical devices).
■​ Soft RTOS: Tolerable delays (e.g., streaming).
4.​ Distributed OS
○​ Manages resources across networked computers.
○​ Example: Google's Borg system.
5.​ Embedded OS
○​ Lightweight, for dedicated hardware (e.g., Android on IoT).

2. CPU Scheduling

Key Concepts

●​ Preemption: OS can interrupt a running process.

●​ Context Switch: Saving/restoring process state.
●​ Scheduling Criteria:
○​ CPU utilization, throughput, turnaround time, waiting time, response
time.

Scheduling Algorithms

Algorithm Description Pros & Cons

FCFS First-Come-First-Served Simple, but poor for short jobs

SJF Shortest Job First Optimal avg wait time, but hard to

predict job length

Priority Assigns priority levels Risk of starvation (low-priority jobs)

Round Robin Time slices (quantum) for Fair, but high context switches

each job

Multilevel Separate queues for different Complex but flexible

Queue job types

3. Concurrent Programming

Key Challenges

1.​ Race Conditions

○​ Occur when multiple processes access shared data concurrently.
○​ Solved via synchronization (e.g., locks, semaphores).
2.​ Critical Sections
○​ Code segments accessing shared resources.
○​ Requirements:
■​ Mutual exclusion.
■​ Progress (no indefinite waiting).
■​ Bounded waiting.

Synchronization Tools

Tool Description Use Case

Semaphores Integer with wait()/signal() Resource counting

Mutex Locks Binary locks (acquire/release) Simple mutual exclusion

Monitors Encapsulates data + sync High-level abstraction

procedures (Java)

Condition wait()/signal() within monitors Coordinating thread

Variables actions
Deadlocks

●​ Definition: Circular waiting for resources.

●​ Necessary Conditions (Coffman):
○​ Mutual Exclusion
○​ Hold and Wait
○​ No Preemption
○​ Circular Wait
●​ Handling Techniques:
○​ Prevention: Break one condition (e.g., request all resources upfront).
○​ Avoidance: Banker’s algorithm (safe state checks).
○​ Detection & Recovery: Kill processes or preempt resources.

Key Comparisons

Concept Batch OS vs. Time-Sharing

Interaction None vs. Multi-user interactive

Efficiency High throughput vs. Fast

response

Algorithm FCFS vs. Round Robin

Fairness Unfair to short jobs vs.

Equitable

Overhead Low vs. High (context

switches)

Memory Management in Operating Systems

Memory management is a critical function of an OS that ensures efficient allocation,

tracking, and optimization of memory usage. It includes both contiguous and

non-contiguous allocation techniques, as well as virtual memory systems.

1. Contiguous Memory Allocation

Memory is allocated in a single continuous block.

A. Static Allocation (Fixed Partitioning)

●​ Memory is divided into fixed-size partitions.

●​ Disadvantages:
○​ Internal Fragmentation: Unused memory within a partition.
○​ Limited flexibility (e.g., a large program may not fit).

B. Dynamic Partitioning (Variable Partitioning)

●​ Memory is allocated in variable-sized blocks as needed.

●​ Algorithms:
○​ First Fit: Allocate the first available block that fits.
○​ Best Fit: Allocate the smallest suitable block.
○​ Worst Fit: Allocate the largest available block (rarely used).
●​ Disadvantages:
○​ External Fragmentation: Free memory becomes scattered.
○​ Compaction (memory shuffling) may be needed.

C. Overlays

●​ Used when program size > available memory.

●​ Only required modules are loaded at runtime.
●​ Disadvantage: Manual management (rarely used today).

2. Non-Contiguous Memory Allocation

Memory is allocated in disjointed blocks, reducing fragmentation.

A. Paging

●​ Divides memory into fixed-size blocks (frames) and programs into pages.
●​ Page Table: Maps logical addresses to physical frames.
●​ Advantages:
○​ No external fragmentation.
○​ Efficient for virtual memory.
●​ Disadvantages:
○​ Internal fragmentation (last page may be partially unused).
○​ Overhead due to page tables.

B. Segmentation

●​ Divides memory into variable-sized segments (e.g., code, data, stack).

●​ Segment Table: Stores base address and segment length.
●​ Advantages:
○​ Matches program structure better than paging.
●​ Disadvantages:
○​ External fragmentation possible.
○​ Slower than paging due to variable sizes.

C. Paged Segmentation

●​ Combines paging and segmentation:

○​ Segments are divided into pages.
○​ Reduces external fragmentation while keeping benefits of
segmentation.

3. Virtual Memory

Allows programs to use more memory than physically available via demand paging.

A. Demand Paging

●​ Pages are loaded into memory only when needed.

●​ Page Fault: Occurs when a required page is not in RAM.
○​ OS fetches it from disk (swap space).

B. Page Replacement Algorithms

When RAM is full, the OS must decide which page to replace:

1.​ FIFO (First-In-First-Out)

○​ Replaces the oldest page.
○​ Disadvantage: May replace frequently used pages.
2.​ LRU (Least Recently Used)
○​ Replaces the least recently accessed page.
○​ Optimal but hard to implement.
3.​ Clock (Second Chance)
○​ Approximates LRU using a reference bit.
4.​ Optimal (Theoretical)
○​ Replaces the page that won’t be used for the longest time (not
practical).

C. Thrashing

●​ Occurs when excessive page faults slow down the system.

●​ Solution:
○​ Increase RAM.
○​ Use better page replacement (e.g., LRU).
○​ Adjust working set (number of active pages per process).

4. Hardware Support for Memory Management

●​ MMU (Memory Management Unit): Translates logical → physical addresses.

●​ TLB (Translation Lookaside Buffer): Cache for page table entries (speeds up
address translation).
●​ Page Table Structures:
○​ Hierarchical (Multi-level): Reduces memory overhead (e.g., 2-level
paging).
○​ Hashed: Used in large address spaces (e.g., Linux).
○​ Inverted: One entry per physical frame (saves space).

Summary Table

Technique Pros Cons

Fixed Partitioning Simple to implement Internal fragmentation

Dynamic Partitioning Flexible allocation External fragmentation

Paging No external fragmentation Internal fragmentation,

overhead

Segmentation Logical program External fragmentation

organization

Demand Paging Efficient memory use Page faults slow performance

LRU Page Minimizes page faults Hard to implement perfectly

Replacement

File Systems in Operating Systems

A file system organizes data storage on disks, providing structured access to files

and directories. It includes:

1.​ Logical File System (user view: files, directories).

2.​ Physical File System (how data is stored on disk).
3.​ Allocation Strategies (how disk blocks are assigned to files).
4.​ Disk Scheduling (optimizing read/write operations).

1. Simple File System Structure

A. File System Components

1.​ File: Named collection of data (text, binary, etc.).

2.​ Directory: Hierarchical structure to organize files.
 3.​ Metadata:
○​ File name, size, permissions, timestamps.
○​ Stored in inodes (UNIX) or file control blocks (FCBs) (Windows).

B. Disk Layout

Section Purpose

Boot Block Stores OS bootloader.

Superblock File system metadata (size, free

blocks).

Inode Table Stores file metadata (UNIX/Linux).

Data Blocks Actual file contents.

2. Physical File System

A. Allocation Strategies

How disk blocks are assigned to files:

(1) Contiguous Allocation

●​ Files occupy sequential disk blocks.

 ●​ Pros:
○​ Fast sequential access.
○​ Simple to implement.
●​ Cons:
○​ External fragmentation (free gaps between files).
○​ Hard to grow files.

(2) Linked Allocation

●​ Files stored as linked lists of blocks.

○​ Each block points to the next.
●​ Pros:
○​ No external fragmentation.
○​ Files can grow dynamically.
●​ Cons:
○​ Slow random access (must traverse links).
○​ Overhead for pointers.

(3) Indexed Allocation

●​ Uses an index block (or inode) to store pointers to all file blocks.
●​ Pros:
○​ Fast random access.
○​ No external fragmentation.
●​ Cons:
○​ Overhead for index blocks.

3. Disk Scheduling Algorithms

Optimize disk head movement to reduce seek time.

Algorithm Description Pros & Cons

FCFS Requests served in Simple, but poor performance.

order.

SSTF (Shortest Seek Picks nearest request. Faster than FCFS, but may

Time First) starve far requests.

SCAN (Elevator) Moves back-and-forth Fair, but long wait at ends.

across disk.

C-SCAN Circular SCAN (resets at More uniform wait times.

end).

LOOK SCAN but reverses Reduces unnecessary

direction early. movement.

4. Device Handlers (Disk Drivers)

●​ Translate OS file requests into low-level disk commands.

●​ Functions:
○​ Read/Write Scheduling: Uses disk scheduling algorithms.
○​ Error Handling: Bad sector recovery.
 ○​ Caching: Stores frequently accessed blocks in RAM (buffer cache).

5. Key File System Examples

File System Features

FAT (File Allocation Simple, used in USB drives. Linked allocation.

Table)

NTFS (Windows) Journaling, security (ACLs), large file support.

ext4 (Linux) Journaling, extents (contiguous blocks), faster

fsck.

ZFS Advanced error correction, snapshots,

RAID-like pooling.

Operating Systems MCQs with Answers

Evolution of Operating Systems

1.​ Which was the first widely used operating system?​

a) UNIX​
b) MS-DOS​
 c) Windows​
d) Linux​
Answer: b) MS-DOS
2.​ Time-sharing systems were developed to:​
a) Improve CPU utilization​
b) Allow multiple users simultaneous access​
c) Handle real-time processes​
d) Manage distributed systems​
Answer: b) Allow multiple users simultaneous access
3.​ Which OS type is used in pacemakers?​
a) Distributed OS​
b) Real-Time OS​
c) Batch OS​
d) Time-Sharing OS​
Answer: b) Real-Time OS

CPU Scheduling

4.​ Which scheduling algorithm always selects the shortest job next?​
a) FCFS​
b) SJF​
c) Round Robin​
d) Priority​
Answer: b) SJF
5.​ The convoy effect occurs in:​
a) Priority Scheduling​
b) FCFS Scheduling​
c) Round Robin​
d) Multilevel Queue​
Answer: b) FCFS Scheduling
6.​ In Round Robin scheduling, if time quantum is too large, it behaves like:​
a) SJF​
b) FCFS​
c) Priority​
d) Multilevel Feedback​
Answer: b) FCFS

Concurrent Programming & Deadlocks

7.​ Which is NOT a necessary condition for deadlock?​

a) Mutual Exclusion​
b) Hold and Wait​
 c) No Preemption​
d) Starvation​
Answer: d) Starvation
8.​ The Banker's Algorithm is used for:​
a) Deadlock Prevention​
b) Deadlock Avoidance​
c) Deadlock Detection​
d) Deadlock Recovery​
Answer: b) Deadlock Avoidance
9.​ A semaphore initialized to 1 is called:​
a) Counting Semaphore​
b) Binary Semaphore​
c) Mutex Lock​
d) Monitor​
Answer: b) Binary Semaphore

Memory Management

10.​Belady's anomaly occurs with which page replacement?​

a) FIFO​
b) LRU​
c) Optimal​
d) Clock​
Answer: a) FIFO
11.​Virtual memory uses:​
a) Only physical memory​
b) Only secondary storage​
c) Both physical memory and disk​
d) Cache memory only​
Answer: c) Both physical memory and disk
12.​The optimal page replacement algorithm:​
a) Removes the least recently used page​
b) Removes the page not used for longest time​
c) Is impossible to implement​
d) Uses a circular buffer​
Answer: b) Removes the page not used for longest time

File Systems

13.​Which disk scheduling algorithm services requests in the direction it's

moving?​
a) FCFS​
 b) SSTF​
c) SCAN​
d) C-LOOK​
Answer: c) SCAN
14.​NTFS uses which allocation method?​
a) Contiguous​
b) Linked​
c) Indexed​
d) Hashed​
Answer: c) Indexed
15.​The file allocation table (FAT) is used in:​
a) Contiguous allocation​
b) Linked allocation​
c) Indexed allocation​
d) Hashed allocation​
Answer: b) Linked allocation

Advanced Topics

16.​Which is true about monitors?​

a) They use message passing​
b) Only one process can execute monitor code at a time​
c) They don't provide synchronization​
d) They are hardware features​
Answer: b) Only one process can execute monitor code at a time
17.​In paging, the page table maps:​
a) Logical to physical addresses​
b) Physical to logical addresses​
c) Files to disk blocks​
d) Processes to CPUs​
Answer: a) Logical to physical addresses
18.​Thrashing occurs when:​
a) CPU utilization is high​
b) Page fault rate is very high​
c) Deadlock occurs​
d) Disk fails​
Answer: b) Page fault rate is very high

Additional Questions

19.​Which scheduling algorithm is most suitable for real-time systems?​

a) FCFS​
 b) SJF​
c) Priority​
d) Round Robin​
Answer: c) Priority
20.​The four necessary conditions for deadlock are known as:​
a) Coffman conditions​
b) Peterson conditions​
c) Dijkstra conditions​
d) Lamport conditions​
Answer: a) Coffman conditions
21.​In segmentation, memory is divided into:​
a) Fixed-size blocks​
b) Variable-size blocks​
c) Equal-size pages​
d) Cache lines​
Answer: b) Variable-size blocks
22.​Which IPC mechanism is fastest?​
a) Pipes​
b) Message queues​
c) Shared memory​
d) Sockets​
Answer: c) Shared memory
23.​The working set model is used for:​
a) Deadlock avoidance​
b) Memory management​
c) Disk scheduling​
d) Process synchronization​
Answer: b) Memory management
24.​Which is NOT a disk scheduling algorithm?​
a) SCAN​
b) SSTF​
c) LOOK​
d) SWAP​
Answer: d) SWAP
25.​In virtual memory, the page fault service time includes:​
a) Only disk I/O time​
b) Only memory access time​
c) Disk I/O + context switch time​
d) CPU execution time​
Answer: c) Disk I/O + context switch time
26.​A race condition occurs when:​
a) Two processes compete for same resource​
 b) Process priority changes dynamically​
c) Memory allocation fails​
d) Disk becomes fragmented​
Answer: a) Two processes compete for same resource
27.​Which page replacement algorithm uses reference bits?​
a) FIFO​
b) LRU​
c) Clock​
d) Optimal​
Answer: c) Clock
28.​The critical section problem requires:​
a) Mutual exclusion​
b) Progress​
c) Bounded waiting​
d) All of the above​
Answer: d) All of the above
29.​Which is true about SJF scheduling?​
a) It can cause starvation​
b) It's always optimal​
c) It's easy to implement​
d) It's non-preemptive​
Answer: a) It can cause starvation
30.​In distributed systems, which protocol is used for mutual exclusion?​
a) Banker's algorithm​
b) Peterson's algorithm​
c) Ricart-Agrawala algorithm​
d) Round Robin algorithm​
Answer: c) Ricart-Agrawala algorithm
31.​The fork() system call creates:​
a) Thread​
b) Process​
c) Pipe​
d) Semaphore​
Answer: b) Process
32.​Which is NOT a component of a file system?​
a) Directory structure​
b) Memory manager​
c) Access methods​
d) Allocation methods​
Answer: b) Memory manager
33.​The dining philosophers problem illustrates:​
a) Memory fragmentation​
 b) Deadlock​
c) Page replacement​
d) Disk scheduling​
Answer: b) Deadlock
34.​Which is true about demand paging?​
a) All pages are loaded at startup​
b) Pages are loaded when needed​
c) It doesn't use disk​
d) It's slower than swapping​
Answer: b) Pages are loaded when needed
35.​The PCB (Process Control Block) contains:​
a) Process state​
b) CPU registers​
c) Memory limits​
d) All of the above​
Answer: d) All of the above
36.​Which is NOT a memory allocation technique?​
a) First-fit​
b) Best-fit​
c) Worst-fit​
d) Quick-fit​
Answer: d) Quick-fit
37.​In UNIX, which system call creates a pipe?​
a) pipe()​
b) fork()​
c) exec()​
d) open()​
Answer: a) pipe()
38.​The principle of locality refers to:​
a) Temporal and spatial locality​
b) Process priority​
c) Deadlock conditions​
d) File allocation​
Answer: a) Temporal and spatial locality
39.​Which is true about multilevel queue scheduling?​
a) Processes can move between queues​
b) Each queue has its own scheduling algorithm​
c) It doesn't use priorities​
d) It's only for real-time systems​
Answer: b) Each queue has its own scheduling algorithm
40.​The purpose of TLB (Translation Lookaside Buffer) is to:​
a) Speed up address translation​
 b) Detect deadlocks​
c) Schedule processes​
d) Manage files​
Answer: a) Speed up address translation

Process Management

41.​What is the main purpose of a Process Control Block (PCB)?​

a) To store temporary files​
b) To hold process execution state and context​
c) To manage disk partitions​
d) To schedule I/O operations​
Answer: b) To hold process execution state and context
42.​A zombie process is:​
a) A terminated process waiting for its parent to read its status​
b) A process consuming excessive CPU​
c) A process blocked on I/O​
d) A process with memory leaks​
Answer: a) A terminated process waiting for its parent to read its status
43.​The fork() system call returns:​
a) 0 to the child, PID of child to parent​
b) 1 to the child, -1 to parent​
c) Same PID to both processes​
d) Random values​
Answer: a) 0 to the child, PID of child to parent

Threads

44.​User-level threads are managed by:​

a) The operating system kernel​
b) A runtime library in user space​
c) Hardware interrupts​
d) Device drivers​
Answer: b) A runtime library in user space
45.​The main advantage of kernel-level threads over user-level threads is:​
a) Lower creation overhead​
b) Better parallelism on multiprocessors​
c) No need for synchronization​
d) Simpler implementation​
Answer: b) Better parallelism on multiprocessors
46.​In a many-to-one threading model:​
a) Many user threads map to one kernel thread​
 b) One user thread maps to many kernel threads​
c) Each user thread has a dedicated kernel thread​
d) Threads are not supported​
Answer: a) Many user threads map to one kernel thread

Synchronization

47.​The test-and-set instruction is used to:​

a) Implement mutual exclusion​
b) Allocate memory​
c) Schedule processes​
d) Manage file systems​
Answer: a) Implement mutual exclusion
48.​A monitor differs from a semaphore in that:​
a) Monitors are language constructs while semaphores are OS primitives​
b) Semaphores are faster​
c) Monitors don't provide mutual exclusion​
d) Semaphores can't cause deadlocks​
Answer: a) Monitors are language constructs while semaphores are OS
primitives
49.​The readers-writers problem favors:​
a) Readers (multiple can read simultaneously)​
b) Writers (exclusive access)​
c) Neither has priority​
d) Depends on implementation​
Answer: d) Depends on implementation

Memory Management

50.​External fragmentation occurs in:​

a) Paging​
b) Segmentation​
c) Virtual memory​
d) Cache memory​
Answer: b) Segmentation
51.​The working set of a process is:​
a) The set of pages currently in physical memory​
b) The set of pages accessed in the last Δ time units​
c) All pages the process will ever need​
d) The pages swapped out to disk​
Answer: b) The set of pages accessed in the last Δ time units
 52.​The purpose of a TLB is to:​
a) Cache page table entries​
b) Manage disk blocks​
c) Schedule processes​
d) Handle interrupts​
Answer: a) Cache page table entries

File Systems

53.​Inode in UNIX systems contains:​

a) File metadata and block pointers​
b) Only file name​
c) Only access permissions​
d) User credentials​
Answer: a) File metadata and block pointers
54.​Journaling in file systems helps with:​
a) Faster file access​
b) Recovery after crashes​
c) Reducing disk space usage​
d) Improving CPU scheduling​
Answer: b) Recovery after crashes
55.​The main advantage of indexed allocation over linked allocation is:​
a) Better random access​
b) Less memory usage​
c) Simpler implementation​
d) No need for disk space​
Answer: a) Better random access

I/O Systems

56.​DMA (Direct Memory Access) is used to:​

a) Bypass CPU for bulk data transfers​
b) Increase CPU clock speed​
c) Manage virtual memory​
d) Schedule processes​
Answer: a) Bypass CPU for bulk data transfers
57.​Which is NOT a disk scheduling algorithm?​
a) SCAN​
b) SSTF​
c) LOOK​
d) SWAP​
Answer: d) SWAP
 58.​Spooling is used for:​
a) Managing print jobs​
b) Memory allocation​
c) Process scheduling​
d) File compression​
Answer: a) Managing print jobs

Security & Protection

59.​The principle of least privilege means:​

a) Processes get only necessary permissions​
b) All processes run as root​
c) Memory is shared freely​
d) No access control is needed​
Answer: a) Processes get only necessary permissions
60.​A capability list contains:​
a) Objects a process can access and permitted operations​
b) Process scheduling information​
c) Memory allocation details​
d) Disk block addresses​
Answer: a) Objects a process can access and permitted operations

Virtualization

61.​Type 1 hypervisor runs:​

a) On bare metal​
b) Within a host OS​
c) Only on cloud systems​
d) Without hardware support​
Answer: a) On bare metal
62.​Para-virtualization differs from full virtualization by:​
a) Requiring modified guest OS​
b) Not needing hardware support​
c) Being slower​
d) Not supporting multiple VMs​
Answer: a) Requiring modified guest OS

Distributed Systems

63.​The CAP theorem states that a distributed system can't simultaneously

guarantee:​
a) Consistency, Availability, Partition tolerance​
 b) Computation, Accuracy, Performance​
c) Concurrency, Access, Persistence​
d) Clock, Accuracy, Precision​
Answer: a) Consistency, Availability, Partition tolerance
64.​Lamport timestamps are used to:​
a) Order events in distributed systems​
b) Schedule CPU processes​
c) Manage memory pages​
d) Allocate disk blocks​
Answer: a) Order events in distributed systems

Real-Time Systems

65.​Hard real-time systems require:​

a) Guaranteed deadline meeting​
b) Best-effort timing​
c) No timing constraints​
d) Only average performance​
Answer: a) Guaranteed deadline meeting
66.​Rate-monotonic scheduling assigns priorities based on:​
a) Task periods (shorter period = higher priority)​
b) Task execution time​
c) Random assignment​
d) Memory requirements​
Answer: a) Task periods (shorter period = higher priority)

Miscellaneous Advanced Topics

67.​The Belady's anomaly can occur with:​

a) FIFO page replacement​
b) LRU page replacement​
c) Optimal page replacement​
d) All page replacement algorithms​
Answer: a) FIFO page replacement
68.​The purpose of the Second Chance page replacement algorithm is to:​
a) Approximate LRU at lower cost​
b) Implement optimal replacement​
c) Eliminate page faults completely​
d) Reduce disk I/O bandwidth​
Answer: a) Approximate LRU at lower cost
69.​The ABA problem occurs in:​
a) Lock-free programming​
 b) Deadlock detection​
c) Memory allocation​
d) File system design​
Answer: a) Lock-free programming
70.​Copy-on-write is used to:​
a) Optimize process forking​
b) Schedule disk I/O​
c) Manage CPU caches​
d) Implement semaphores​
Answer: a) Optimize process forking

Linux/UNIX Specific

71.​In Linux, the init process typically has PID:​

a) 0​
b) 1​
c) 255​
d) 1024​
Answer: b) 1
72.​The 'nice' command affects:​
a) Process priority​
b) Memory allocation​
c) File permissions​
d) Disk scheduling​
Answer: a) Process priority

Windows Specific

73.​The Windows Executive runs in:​

a) Kernel mode​
b) User mode​
c) Both modes​
d) Hardware directly​
Answer: a) Kernel mode
74.​NTFS uses which technique for small files?​
a) Resident attributes​
b) Journaling​
c) RAID​
d) Virtual memory​
Answer: a) Resident attributes

Cloud Computing
 75.​Live migration of VMs requires:​
a) Memory and state transfer without downtime​
b) Complete VM restart​
c) Changing hypervisor type​
d) Physical server replacement​
Answer: a) Memory and state transfer without downtime
76.​Containerization differs from virtualization by:​
a) Sharing host OS kernel​
b) Requiring more resources​
c) Needing special hardware​
d) Being slower​
Answer: a) Sharing host OS kernel

Performance

77.​Thrashing can be detected by:​

a) High page fault rate​
b) Low CPU utilization​
c) Both a and b​
d) Neither a nor b​
Answer: c) Both a and b
78.​The working set model helps prevent:​
a) Thrashing​
b) Deadlocks​
c) Race conditions​
d) Memory leaks​
Answer: a) Thrashing

Emerging Trends

79.​Persistent memory (PMEM) differs from traditional RAM by:​

a) Retaining data after power loss​
b) Being slower​
c) Volatility​
d) Smaller capacity​
Answer: a) Retaining data after power loss
80.​Microkernels differ from monolithic kernels by:​
a) Moving most functionality to user space​
b) Being larger​
c) Having no security features​
d) Not supporting processes​
Answer: a) Moving most functionality to user space
Advanced Memory Management

81.​What is the primary purpose of memory-mapped files?​

a) To allow files to be accessed as memory addresses​
b) To increase disk storage capacity​
c) To replace virtual memory​
d) To speed up CPU cache​
Answer: a) To allow files to be accessed as memory addresses
82.​The slab allocation technique is primarily used for:​
a) Kernel object allocation​
b) User process memory​
c) Disk block management​
d) Network packet buffering​
Answer: a) Kernel object allocation

File Systems & Storage

83.​ZFS (Zettabyte File System) uses which innovative feature?​

a) Copy-on-write transactional model​
b) Fixed-size inodes​
c) No checksumming​
d) Single-disk reliability​
Answer: a) Copy-on-write transactional model
84.​RAID 5 provides:​
a) Striping with distributed parity​
b) Mirroring without parity​
c) Double mirroring​
d) No fault tolerance​
Answer: a) Striping with distributed parity

Virtualization & Cloud

85.​KVM (Kernel-based Virtual Machine) converts Linux into:​

a) A Type-1 hypervisor​
b) A Type-2 hypervisor​
c) A container runtime​
d) A microkernel​
Answer: a) A Type-1 hypervisor
86.​Docker uses which Linux feature for isolation?​
a) cgroups and namespaces​
b) Kernel modules​
c) System calls​
 d) Interrupt handlers​
Answer: a) cgroups and namespaces

Security & Protection

87.​ASLR (Address Space Layout Randomization) protects against:​

a) Buffer overflow attacks​
b) Deadlocks​
c) Disk failures​
d) Memory leaks​
Answer: a) Buffer overflow attacks
88.​SELinux implements:​
a) Mandatory Access Control​
b) Discretionary Access Control​
c) No access control​
d) Only file permissions​
Answer: a) Mandatory Access Control

Real-Time Systems

89.​In priority inheritance protocol:​

a) A lower priority task inherits higher priority when blocking a high-priority
task​
b) Priorities remain fixed​
c) All tasks get equal priority​
d) Priorities decrease over time​
Answer: a) A lower priority task inherits higher priority when blocking a
high-priority task
90.​Earliest Deadline First (EDF) scheduling is:​
a) Dynamic priority scheduling​
b) Fixed priority scheduling​
c) Non-preemptive​
d) Only for non-real-time systems​
Answer: a) Dynamic priority scheduling

Distributed Systems

91.​The Byzantine Generals Problem relates to:​

a) Fault tolerance in distributed consensus​
b) Memory allocation​
c) Process scheduling​
 d) File system design​
Answer: a) Fault tolerance in distributed consensus
92.​Paxos algorithm is used for:​
a) Distributed consensus​
b) Memory management​
c) Disk scheduling​
d) Process synchronization​
Answer: a) Distributed consensus

Performance & Optimization

93.​The Bélády anomaly demonstrates that:​

a) More page frames can sometimes increase page faults​
b) LRU is always optimal​
c) Memory is unlimited​
d) Disk I/O is unimportant​
Answer: a) More page frames can sometimes increase page faults
94.​The working set model helps determine:​
a) How much memory a process needs​
b) CPU clock speed​
c) Disk rotation speed​
d) Network bandwidth​
Answer: a) How much memory a process needs

Contemporary Topics

95.​Non-Uniform Memory Access (NUMA) architectures:​

a) Have different memory access times depending on location​
b) Provide uniform memory access times​
c) Eliminate memory entirely​
d) Only use disk storage​
Answer: a) Have different memory access times depending on location
96.​Persistent memory (e.g., Intel Optane) blurs the line between:​
a) Memory and storage​
b) CPU and GPU​
c) Kernel and user space​
d) Processes and threads​
Answer: a) Memory and storage

Linux/Windows Internals
 97.​The Linux OOM (Out-Of-Memory) killer:​
a) Selectively terminates processes to free memory​
b) Crashes the entire system​
c) Expands swap space automatically​
d) Slows down all processes equally​
Answer: a) Selectively terminates processes to free memory
98.​Windows NT's Object Manager provides:​
a) Uniform interface for all system resources​
b) Only file management​
c) Just memory management​
d) Exclusive CPU scheduling​
Answer: a) Uniform interface for all system resources

Emerging Technologies

99.​Unikernels differ from traditional OS by:​

a) Including only necessary OS components for each application​
b) Being larger than monolithic kernels​
c) Requiring hardware virtualization​
d) Not supporting networking​
Answer: a) Including only necessary OS components for each application
100.​ Serverless computing (e.g., AWS Lambda) actually uses:​
a) Short-lived containers behind the scenes​
b) No servers at all​
c) Only physical servers​
d) No operating system​
Answer: a) Short-lived containers behind the scenes