D686 Operating Systems for Computer Scientists - Set 2 - Part 1

Correct Answer: Hypervisor virtualization

Hypervisor virtualization enables multiple operating systems to run simultaneously on single hardware, providing isolation and efficient resource sharing for cloud environments.
Option a is incorrect because monolithic kernel runs single operating system instance with all services in kernel space, not supporting multiple OS instances.
Option b is incorrect because microkernel architecture minimizes kernel services but doesn't inherently support multiple operating system instances on same hardware platform.
Option d is incorrect because layered architecture organizes single OS into hierarchical levels, not enabling multiple isolated operating systems to run concurrently.

Correct Answer: Boot loader

Boot loader executes after BIOS/UEFI, initializing system and loading kernel from storage into memory to begin operating system startup sequence properly.
Option a is incorrect because device drivers load after kernel initialization to provide hardware interfaces, not during the initial boot loading sequence.
Option c is incorrect because shell programs provide user interfaces after the operating system fully loads, not during early boot processes fundamentally.
Option d is incorrect because system utilities run after OS initialization to provide services, not handling the initial kernel loading during boot.

Correct Answer: User-level threading

User-level threading manages threads in user space without kernel involvement, reducing overhead and context switch costs crucial for resource-constrained IoT devices.
Option a is incorrect because kernel-level threading requires kernel resources for each thread, increasing overhead unacceptable for limited embedded system resources available.
Option c is incorrect because hybrid threading combines both models adding complexity inappropriate for simple IoT devices with minimal resources and requirements.
Option d is incorrect because hardware threading requires specialized processor support not typically available in low-cost IoT embedded system processors generally.

Correct Answer: Longer time quantum

Longer time quantum reduces context switch frequency by allowing threads to run longer before switching, improving throughput while maintaining reasonable responsiveness.
Option a is incorrect because frequent time slices increase context switches, worsening the performance problem rather than solving the excessive switching issue.
Option c is incorrect because immediate preemption maximizes context switches for responsiveness, contradicting the goal of reducing switching overhead for performance.
Option d is incorrect because random scheduling provides no optimization for context switch reduction and may worsen performance through unpredictable behavior patterns.

Correct Answer: Background services

Background services run independently of user interface, enabling continuous operation for tasks like audio playback when apps aren't actively displayed.
Option a is incorrect because foreground services show persistent notifications and run while app is visible, not specifically for background operation continuation.
Option c is incorrect because system processes handle OS-level operations, not application-specific background tasks like media playback in user applications typically.
Option d is incorrect because kernel threads execute system-level code in kernel space, not application-level services like music playback in user apps.

Correct Answer: Direct memory access

Memory-mapped files appear in process address space, allowing direct memory operations without system calls, reducing overhead for frequent file access patterns.
Option a is incorrect because memory mapping doesn't change file size; it changes how files are accessed by mapping them into memory.
Option c is incorrect because memory mapping doesn't provide encryption; it's an access method independent of security features like encryption mechanisms.
Option d is incorrect because memory mapping works with local files in virtual memory, not providing network file access transparency automatically.

Correct Answer: Network time protocol

NTP synchronizes clocks across networked systems, maintaining time consistency essential for distributed transaction ordering and coordinated operations between servers effectively.
Option a is incorrect because local timers drift independently on each system, causing time inconsistencies that break distributed transaction ordering requirements critically.
Option c is incorrect because CPU cycles vary with processor speed and load, unsuitable for consistent time measurement across different server hardware.
Option d is incorrect because interrupt counters track system events not time, failing to provide synchronized timestamps needed for distributed systems.

Correct Answer: Journaling capability

Journaling file systems log changes before committing them, enabling quick recovery by replaying journal rather than checking entire file system structure.
Option a is incorrect because compression reduces storage space usage but doesn't address consistency checking or recovery time after system crashes occur.
Option c is incorrect because encryption protects data confidentiality, not improving file system recovery or consistency checking performance after improper shutdowns happen.
Option d is incorrect because quota management controls user space allocation, unrelated to crash recovery or consistency checking optimization for file systems.

Correct Answer: Device drivers

Device drivers translate between standard kernel interfaces and specific hardware implementations, abstracting device differences while providing consistent kernel interaction methods effectively.
Option a is incorrect because system calls provide user-to-kernel interface for services, not hardware abstraction for different device types and implementations.
Option c is incorrect because boot firmware initializes hardware during startup but doesn't provide ongoing abstraction for device variety during operation.
Option d is incorrect because user applications consume OS services through APIs, not providing hardware abstraction layer between kernel and devices.

Correct Answer: Mode switching

Mode switching transitions between user and kernel modes through controlled entry points, enabling temporary privilege elevation for system operations securely.
Option b is incorrect because context switching changes between processes or threads, preserving privilege level rather than transitioning between protection domains.
Option c is incorrect because thread switching alternates between threads within same privilege level, not changing protection domain for privileged operations.
Option d is incorrect because process migration moves processes between processors or systems, unrelated to protection domain transitions for privilege changes.

Correct Answer: Deadlock exists

Cycles in single-instance resource graphs definitively indicate deadlock since processes form circular wait chains that cannot be resolved without intervention.
Option a is incorrect because safe states have no cycles, guaranteeing all processes can complete by careful resource allocation ordering and management.
Option c is incorrect because starvation involves indefinite waiting without circular dependencies; cycles specifically indicate deadlock not just starvation conditions present.
Option d is incorrect because while preemption might resolve deadlocks, cycles indicate deadlock existence rather than suggesting preemption as the solution method.

Correct Answer: Hardware virtualization

Hardware virtualization extensions trap and redirect privileged instructions efficiently, enabling near-native performance for guest operating systems without software interpretation overhead.
Option a is incorrect because cache memory improves data access speed but doesn't specifically handle privileged instruction execution for virtual machines.
Option c is incorrect because pipeline processing improves instruction throughput generally but doesn't address virtual machine privileged instruction execution specifically needed.
Option d is incorrect because branch prediction optimizes conditional execution flow, not providing privileged instruction handling for virtualized operating systems fundamentally.

Correct Answer: Memory pressure

Memory pressure triggers swapping when physical memory becomes scarce, moving less-used pages to disk to free RAM for active processes.
Option a is incorrect because system startup loads essential components into memory; swapping occurs during runtime when memory becomes insufficient for demands.
Option c is incorrect because file creation affects disk space not memory pressure directly, though large files might indirectly cause memory pressure.
Option d is incorrect because network activity uses buffers but doesn't directly trigger swapping unless it causes overall memory shortage conditions.

Correct Answer: Default action occurs

Unhandled signals trigger default actions defined by the operating system, which may include termination, ignoring, or core dump generation depending on signal type.
Option a is incorrect because signals interrupt normal execution flow; ignoring them completely would defeat their purpose for asynchronous event notification fundamentally.
Option c is incorrect because receiving signals doesn't automatically cause waiting state; the response depends on signal type and handler configuration present.
Option d is incorrect because signal reception doesn't create child processes; that requires explicit fork calls not automatic signal response behavior.

Correct Answer: Cache coherence

Cache coherence problems arise when processors cache shared data independently, requiring protocols to maintain consistency across all cached copies in multiprocessor systems.
Option a is incorrect because cache miss occurs when requested data isn't in cache, different from inconsistency between multiple cached copies problem.
Option c is incorrect because cache overflow isn't standard terminology; caches use replacement policies when full rather than overflowing like buffers.
Option d is incorrect because cache fragmentation doesn't occur since caches use fixed-size lines, unlike memory allocation that can fragment over time.

Correct Answer: Page fault rate

High page fault rates indicate frequent swapping due to insufficient memory, directly measuring memory pressure and potential thrashing conditions affecting performance.
Option a is incorrect because CPU utilization shows processor load but doesn't directly indicate memory sufficiency or swapping activity in the system.
Option c is incorrect because network throughput measures data transfer rates over network connections, unrelated to memory adequacy for current workload requirements.
Option d is incorrect because disk capacity shows storage availability not memory sufficiency; full disks affect swap space not RAM adequacy.

Correct Answer: Race conditions

Race conditions occur when processors concurrently modify shared kernel structures without synchronization, causing data corruption and system instability requiring careful locking.
Option a is incorrect because power consumption is hardware concern not directly related to concurrent kernel code execution synchronization issues in SMP.
Option c is incorrect because heat generation is physical issue from processor activity, not the software synchronization problem of concurrent kernel access.
Option d is incorrect because clock skew involves timing differences between processors, not the data consistency issues from unsynchronized kernel modifications.

Correct Answer: No user interaction

Batch systems process jobs without user interaction after submission, optimizing for throughput rather than response time unlike interactive systems requiring user feedback.
Option a is incorrect because batch systems don't provide real-time response; they focus on throughput over latency for non-interactive job processing.
Option c is incorrect because interface type doesn't define batch processing; batch jobs can be submitted through various interfaces including graphical ones.
Option d is incorrect because network connectivity is independent of batch versus interactive distinction; both system types can utilize network resources.

Correct Answer: Hardware interrupts

Hardware interrupts signal the CPU asynchronously when devices need attention, avoiding wasteful polling while enabling immediate response to device events.
Option a is incorrect because polling loops waste CPU cycles continuously checking device status rather than being notified when attention is needed.
Option c is incorrect because shared memory provides data exchange but doesn't actively notify the processor when device operations complete or need attention.
Option d is incorrect because system calls are software-initiated requests from programs to kernel, not hardware notifications from device controllers to CPU.

Correct Answer: Reduced creation overhead

Thread pools eliminate thread creation and destruction costs by reusing existing threads, significantly improving response time for request handling in servers.
Option a is incorrect because thread pools share process memory space like any threads; they don't provide additional memory isolation between requests.
Option c is incorrect because thread pools add complexity through pool management, work queues, and thread lifecycle handling compared to simple creation.
Option d is incorrect because thread pools don't inherently improve security; threads in pools have same access rights as individually created threads.