Question 1

One of the earliest proposed approaches to providing distributed mutual exclusion is based on the concept of a distributed queue.

Accepted Answer

The concept of a distributed queue was one of the earliest proposed approaches to providing distributed mutual exclusion.

Question 2

The operating system knows the current state of all processes in the distributed system.

Accepted Answer

The operating system does not necessarily know the current state of all processes in a distributed system as each process may be running on a different computer and the operating system only has direct control over the computer it is running on. However, there may be other mechanisms in place, such as inter-process communication, that allow for some level of visibility and control over the state of processes in a distributed system.

Question 3

Mutual exclusion, deadlock, and starvation are not issues that are faced in a distributed system.

Accepted Answer

Mutual exclusion, deadlock, and starvation are still potential issues in distributed systems. Distributed systems have multiple nodes that may be accessing the same resource or data, which can result in race conditions, deadlocks, and starvation if proper synchronization mechanisms are not implemented. Additionally, network latency and message delays in distributed systems can exacerbate these problems.

Question 4

Algorithms for mutual exclusion and deadlock must depend on access to common memory.

Accepted Answer

Mutual exclusion and deadlock algorithms can be implemented in distributed systems where processes do not have access to a common memory but communicate through messages.

Question 5

A __________ process transfer involves only processes that have not begun execution and hence do not require transferring the state of the process.
A) channel
B) global
C) nonpreemptive
D) preemptive

Accepted Answer

A nonpreemptive process transfer involves only processes that have not begun execution and hence there is no need to transfer the state of the process. Nonpreemptive scheduling involves processes voluntarily relinquishing the CPU after completing their current CPU burst, or when initiating an Input/Output operation that may take some time. This type of scheduling avoids context switching overhead but may result in poor average response time.

Question 6

A _________ exists between two processes if they exchange messages.
A) snapshot
B) state
C) global state
D) channel

Accepted Answer

A channel is a communication mechanism that allows two processes to exchange messages with each other, hence a channel is the perfect choice for this definition.

Question 7

An advantage to nonpreemptive process migration is that it reacts well to sudden changes in load distribution.

Accepted Answer

Nonpreemptive process migration does not react well to sudden changes in load distribution because it only migrates processes at their completion or at specific checkpoints, rather than dynamically adjusting to current load conditions.

Question 8

The negotiation mechanism makes sure that the destination system does not refuse to accept the migration of a process to itself.

Accepted Answer

The negotiation mechanism ensures that both the source and destination systems agree on the migration, but it does not guarantee that the destination system will not refuse the migration. The destination system may still refuse the migration based on its own policies, resources, or other constraints.

Question 9

_________ is the combined state of all processes.
A) Channel
B) Global state
C) Snapshot
D) Distributed snapshot

Accepted Answer

The combined state of all processes is referred to as global state. Channels represent the communication links between processes, snapshots are consistent and coordinated captures of the distributed state, and a distributed snapshot is a coordinated capture of all the processes in the distributed system.

Question 10

With the __________ strategy, the entire address space at the time of migration is transferred
A) precopy
B) flushing
C) eager (dirty)
D) eager (all)

Accepted Answer

Eager (all) strategy transfers the entire memory contents at the time of migration to the destination node, regardless of whether the pages are dirty or clean. This strategy ensures low downtime during migration as all the memory pages are transferred in one go.

Question 11

The decision to migrate must be reached jointly by two Starter processes, one on the source node and one on the destination node.

Accepted Answer

The decision to migrate must be reached jointly by two Starter processes, one on the source node and one on the destination node.

Question 12

Interest in the concept of process migration grew out of research into methods of load balancing across multiple networked systems.

Accepted Answer

Process migration was developed as a technique for balancing the workload across multiple networked systems.

Question 13

Caching is not permitted in the Sprite system.

Accepted Answer

Caching is permitted and utilized in the Sprite system to improve performance by reducing disk access times.

Question 14

The __________ strategy transfers only those pages of the address space that are in main memory and have been modified.
A) eager (dirty)
B) precopy
C) copy-on-reference
D) eager (all)

Accepted Answer

The eager (dirty) strategy specifically transfers only the modified pages (dirty pages) of the address space that are in main memory, optimizing the transfer process by not moving unmodified or unused pages.

Question 15

When a process is migrated, it is necessary to destroy the process on the source system and create it on the target system.

Accepted Answer

When a process is migrated, it is moved from the source system to the target system. Therefore, it is necessary to destroy the process on the source system and create it on the target system to ensure that the process continues to run on the target system.

Question 16

It is the delay in communication, coupled with the lack of a common clock, which makes it much more difficult to develop mutual exclusion mechanisms in a distributed system compared to a centralized system.

Accepted Answer

Distributed systems have no centralized clock or memory, so communication between nodes can be delayed, making it difficult to synchronize and coordinate access to shared resources. This makes developing mutual exclusion mechanisms more challenging than in a centralized system where all nodes have access to a shared clock and memory.

Question 17

The aim of an algorithm that deals with deadlock is either to prevent the formation of a circular wait or to detect its actual or potential occurrence.

Accepted Answer

This statement is true. The two common approaches for dealing with deadlock are prevention and detection. Prevention aims to stop circular wait from happening in the first place, while detection involves identifying when deadlock has occurred or is likely to occur.

Question 18

Deadlock handling is not complicated in a distributed system because each node has accurate knowledge of the current state of the overall system and every message transfer between processes involves predictable delays.

Accepted Answer

Deadlock handling can be more complicated in a distributed system because nodes may not have accurate global knowledge of the system, and message transfers may not be predictable due to varying network delays.

Question 19

A distributed snapshot algorithm assumes that messages are delivered in the order that they are sent and that no messages are lost.

Accepted Answer

This is because the distributed snapshot algorithm relies on the assumption that the messages are received in the correct order and without any loss. If there is any loss or delay in the delivery of messages, the algorithm may not function as expected.

Question 20

A true global state cannot be determined because of the time lapse associated with message transfer.

Accepted Answer

Due to the time lapse involved in message transfer, it is not possible to determine a true global state.

Quiz 18: Distributed Process Management

One of the earliest proposed approaches to providing distributed mutual exclusion is based on the concept of a distributed queue.

The operating system knows the current state of all processes in the distributed system.

Mutual exclusion, deadlock, and starvation are not issues that are faced in a distributed system.

Algorithms for mutual exclusion and deadlock must depend on access to common memory.

A __________ process transfer involves only processes that have not begun execution and hence do not require transferring the state of the process.

A _________ exists between two processes if they exchange messages.

An advantage to nonpreemptive process migration is that it reacts well to sudden changes in load distribution.

The negotiation mechanism makes sure that the destination system does not refuse to accept the migration of a process to itself.

_________ is the combined state of all processes.

With the __________ strategy, the entire address space at the time of migration is transferred

The decision to migrate must be reached jointly by two Starter processes, one on the source node and one on the destination node.

Interest in the concept of process migration grew out of research into methods of load balancing across multiple networked systems.

Caching is not permitted in the Sprite system.

The __________ strategy transfers only those pages of the address space that are in main memory and have been modified.

When a process is migrated, it is necessary to destroy the process on the source system and create it on the target system.

It is the delay in communication, coupled with the lack of a common clock, which makes it much more difficult to develop mutual exclusion mechanisms in a distributed system compared to a centralized system.

The aim of an algorithm that deals with deadlock is either to prevent the formation of a circular wait or to detect its actual or potential occurrence.

Deadlock handling is not complicated in a distributed system because each node has accurate knowledge of the current state of the overall system and every message transfer between processes involves predictable delays.

A distributed snapshot algorithm assumes that messages are delivered in the order that they are sent and that no messages are lost.

A true global state cannot be determined because of the time lapse associated with message transfer.