IncludeHelp_logo
Skip Navigation Links
×

Trending Technologies MCQs

Home » Trending Technologies MCQs

Distributed Computing System MCQs (Multiple-Choice Questions)

Practice Distributed Computing System MCQs to test your understanding of distributed architectures, communication protocols, consistency models, fault tolerance, distributed algorithms, synchronization, scalability, and other essential concepts. These multiple-choice questions are designed for students, IT professionals, and learners preparing for computer science examinations and technical interviews.

Distributed Computing System MCQs

A distributed computing system consists of multiple independent computers that communicate and coordinate their activities to perform tasks. These computers work together by exchanging messages, sharing resources, and processing data across a network. Distributed systems are widely used in cloud computing, distributed databases, web applications, scientific computing, and large-scale data processing.

Practicing these Distributed Computing System MCQs helps you understand important concepts such as client-server architecture, peer-to-peer systems, distributed transactions, replication, consensus algorithms, load balancing, and fault recovery. Each question includes four answer options, the correct answer, and an explanation to support learning.

List of Distributed Computing System MCQs

The following questions cover fundamental and advanced topics in distributed computing systems, including architecture, communication, synchronization, consistency, distributed file systems, scalability, security, and practical problem-solving scenarios.

1. What is a distributed computing system?

  1. A single computer that performs all processing tasks
  2. A collection of independent computers that coordinate to perform tasks
  3. A computer system that operates without an operating system
  4. A system that stores all data on a single local disk

Answer: B) A collection of independent computers that coordinate to perform tasks

Explanation:

A distributed computing system consists of multiple independent computers connected through a network. They communicate and coordinate their operations to accomplish tasks that may involve shared data or computational resources.

2. Which of the following is a major characteristic of a distributed system?

  1. All components must share the same physical memory
  2. Only one processor can execute tasks
  3. Components communicate through a network
  4. All computers must use identical hardware

Answer: C) Components communicate through a network

Explanation:

Distributed systems consist of separate computers that exchange information over a communication network. Their processors and memories are generally independent, and the computers do not need identical hardware.

3. Which architecture organizes a distributed application into clients and servers?

  1. Client-server architecture
  2. Single-user architecture
  3. Standalone architecture
  4. Monolithic hardware architecture

Answer: A) Client-server architecture

Explanation:

In client-server architecture, clients request services or resources, while servers process those requests and return responses. Web applications commonly use this architecture for communication between browsers and application servers.

4. What is the primary purpose of a peer-to-peer (P2P) architecture?

  1. To require every request to pass through one central server
  2. To prevent computers from sharing resources
  3. To store every file on a single computer
  4. To allow participating nodes to share resources directly

Answer: D) To allow participating nodes to share resources directly

Explanation:

In a peer-to-peer system, participating computers can act as both resource providers and consumers. This reduces dependence on a single central server, although coordination and resource discovery may still require specialized mechanisms.

5. What is a node in a distributed computing system?

  1. A communication error only
  2. An individual computer or computing entity participating in the system
  3. A type of database query
  4. A physical cable used for networking

Answer: B) An individual computer or computing entity participating in the system

Explanation:

A node is a participating entity in a distributed system. It may be a physical machine, virtual machine, server, container, or other computing component that performs processing or provides services.

6. Which communication method is commonly used by distributed processes on different machines?

  1. Direct access to the same CPU registers
  2. Shared local processor cache
  3. Message passing
  4. Accessing another machine's physical RAM without communication mechanisms

Answer: C) Message passing

Explanation:

Message passing allows processes on different machines to exchange data and coordinate their actions. Messages may be sent through protocols such as TCP, UDP, or application-level messaging systems.

7. What does scalability mean in a distributed system?

  1. The ability to handle increased workload or system size by adding resources or improving capacity
  2. The ability to operate only on one computer
  3. The requirement to reduce the number of users over time
  4. The elimination of all network communication

Answer: A) The ability to handle increased workload or system size by adding resources or improving capacity

Explanation:

Scalability describes how well a system accommodates growth in users, data, requests, or participating nodes. A scalable system can expand its capacity while maintaining acceptable performance and reliability.

8. What is fault tolerance in distributed computing?

  1. The ability to avoid using networks
  2. The ability to prevent every possible hardware failure
  3. The ability to run only when all nodes are available
  4. The ability to continue providing specified services despite certain component failures

Answer: D) The ability to continue providing specified services despite certain component failures

Explanation:

Fault tolerance allows a distributed system to continue operating when components fail. Techniques such as replication, redundancy, failover, and recovery mechanisms help reduce the impact of failures.

9. Which technique stores copies of data on multiple nodes to improve availability?

  1. Serialization
  2. Replication
  3. Compilation
  4. Fragmentation of source code

Answer: B) Replication

Explanation:

Replication maintains multiple copies of data or services across different nodes. If one replica becomes unavailable, another may continue serving requests, provided the replication design and failure conditions support it.

10. Which of the following is a common challenge in distributed systems?

  1. Every process automatically shares one global clock
  2. Network delays are always constant
  3. Coordinating independent nodes in the presence of communication delays and failures
  4. All machines always execute instructions at the same speed

Answer: C) Coordinating independent nodes in the presence of communication delays and failures

Explanation:

Distributed systems must coordinate components that operate independently and communicate over networks. Variable latency, message loss, node failures, and the absence of a perfectly synchronized global clock make coordination difficult.

11. What is the primary purpose of Remote Procedure Call (RPC)?

  1. To allow a program to invoke an operation on another system using a procedure-call abstraction
  2. To physically connect two processors
  3. To eliminate the need for network protocols
  4. To guarantee that every remote operation succeeds

Answer: A) To allow a program to invoke an operation on another system using a procedure-call abstraction

Explanation:

RPC allows a program to request execution of a procedure on a remote machine. The RPC framework handles communication details such as request encoding, transmission, and response processing, but network failures and timeouts can still occur.

12. Which protocol provides reliable, ordered byte-stream communication between endpoints?

  1. ARP
  2. ICMP
  3. DNS
  4. TCP

Answer: D) TCP

Explanation:

Transmission Control Protocol (TCP) provides a reliable, ordered byte stream between communicating endpoints. Applications must still define their own message boundaries and handle application-level failures, timeouts, and retries.

13. What is middleware in a distributed system?

  1. A hardware component that replaces every server
  2. Software that provides common services and communication support between distributed components
  3. A physical network connection only
  4. A method of disabling remote access

Answer: B) Software that provides common services and communication support between distributed components

Explanation:

Middleware sits between applications and underlying infrastructure. It may provide messaging, service discovery, authentication, serialization, transaction coordination, and other common distributed computing capabilities.

14. Which statement best describes horizontal scaling?

  1. Increasing the clock speed of one processor only
  2. Adding more memory to one machine only
  3. Adding more machines or nodes to share the workload
  4. Reducing the number of application instances

Answer: C) Adding more machines or nodes to share the workload

Explanation:

Horizontal scaling increases capacity by adding more machines or instances. Vertical scaling, in contrast, increases the resources of an existing machine, such as CPU, RAM, or storage capacity.

15. What is load balancing in distributed computing?

  1. Distributing requests or workloads across multiple available resources
  2. Sending every request to a permanently fixed server regardless of its condition
  3. Removing all backup servers
  4. Preventing clients from accessing services

Answer: A) Distributing requests or workloads across multiple available resources

Explanation:

Load balancing distributes incoming requests or computational tasks among multiple servers or nodes. It can improve resource utilization, response time, and availability when combined with appropriate health checks and routing policies.

16. What is the main role of a distributed naming service?

  1. To execute all application code centrally
  2. To synchronize every processor instruction
  3. To prevent machines from communicating
  4. To map names to relevant resources, addresses, or service information

Answer: D) To map names to relevant resources, addresses, or service information

Explanation:

A naming service helps clients locate resources using meaningful names rather than needing to know their physical locations. DNS and service discovery systems are examples of mechanisms that help resolve names to network destinations or service instances.

17. Why is clock synchronization difficult in distributed systems?

  1. Every computer uses the same physical clock
  2. Each machine has its own clock, which may drift and differ from other clocks
  3. Network messages always arrive simultaneously
  4. Distributed systems do not use timestamps

Answer: B) Each machine has its own clock, which may drift and differ from other clocks

Explanation:

Physical clocks on separate machines may run at slightly different rates and start with different offsets. Network delays also complicate synchronization. Protocols such as NTP help synchronize clocks, while logical clocks help represent event ordering without requiring perfectly synchronized physical clocks.

18. What is a logical clock used for in a distributed system?

  1. To measure the exact temperature of a server
  2. To guarantee identical physical timestamps on every machine
  3. To represent causal or ordering relationships between events
  4. To increase processor clock frequency

Answer: C) To represent causal or ordering relationships between events

Explanation:

Logical clocks, such as Lamport clocks, assign logical timestamps to events. They help reason about event ordering and causality even when machines do not share a perfectly synchronized physical clock.

19. What does the term concurrency mean in distributed computing?

  1. Multiple operations can be in progress during overlapping periods
  2. Only one process can ever exist in the system
  3. All tasks must execute sequentially on one CPU
  4. Network communication is disabled during processing

Answer: A) Multiple operations can be in progress during overlapping periods

Explanation:

Concurrency occurs when multiple operations are in progress at overlapping times. In distributed systems, operations may run on separate machines, creating coordination challenges when they access shared data or depend on one another.

20. What is a distributed deadlock?

  1. A temporary increase in network bandwidth
  2. A situation in which all nodes automatically restart
  3. A condition in which a server has no active users
  4. A situation in which distributed processes wait indefinitely for resources held by one another

Answer: D) A situation in which distributed processes wait indefinitely for resources held by one another

Explanation:

A distributed deadlock can occur when processes on different machines form a cycle of resource dependencies and each waits for another process to release a resource. Detecting it can be difficult because dependency information is spread across multiple nodes.

21. What does consistency mean in the context of replicated data?

  1. Every replica must use a different data format
  2. Replica values and observations must obey the system's defined consistency guarantees
  3. Every read must always access the same physical machine
  4. Data must never be updated

Answer: B) Replica values and observations must obey the system's defined consistency guarantees

Explanation:

Consistency defines the rules governing what clients can observe when data is accessed or updated across replicas. Strong consistency models provide stricter guarantees than eventual consistency, which permits temporary differences while updates propagate.

22. What is eventual consistency?

  1. A guarantee that all reads always return the latest globally committed value
  2. A rule that prohibits data replication
  3. A model in which replicas converge if updates stop and the system can continue propagating them
  4. A mechanism that eliminates network delays

Answer: C) A model in which replicas converge if updates stop and the system can continue propagating them

Explanation:

Eventual consistency allows replicas to temporarily contain different values. If updates cease and propagation continues successfully, replicas are expected to converge. Applications must account for potentially stale reads during the convergence period.

23. What is the CAP theorem commonly used to explain?

  1. The trade-off between consistency and availability when a network partition occurs
  2. The relationship between CPU temperature and memory size
  3. The need to use only one server in a database
  4. The elimination of all communication failures

Answer: A) The trade-off between consistency and availability when a network partition occurs

Explanation:

The CAP theorem states that, during a network partition, a distributed data system cannot guarantee both strong consistency and availability for every request. The system must make trade-offs in how it responds to requests affected by the partition. Partition tolerance is essential when network partitions must be handled.

24. What is a network partition in a distributed system?

  1. A process of dividing source code into functions
  2. A method of increasing the capacity of a CPU
  3. A technique for compressing files
  4. A communication failure that separates nodes into groups unable to communicate normally across the affected network paths

Answer: D) A communication failure that separates nodes into groups unable to communicate normally across the affected network paths

Explanation:

A network partition occurs when some nodes cannot communicate with other nodes because of a network failure. Nodes on each side may remain operational, making it necessary to decide whether to serve requests, reject them, or restrict certain operations.

25. What is the primary purpose of a consensus algorithm?

  1. To ensure every node has identical hardware
  2. To enable participating nodes to agree on a value or sequence of decisions under specified failure assumptions
  3. To eliminate the need for communication
  4. To make every network request execute instantly

Answer: B) To enable participating nodes to agree on a value or sequence of decisions under specified failure assumptions

Explanation:

Consensus algorithms help distributed nodes agree on decisions despite certain failures and communication delays. They are used in replicated logs, coordination services, and distributed databases. Their guarantees depend on the algorithm's assumptions about failures and network behavior.

26. Which of the following is a consensus algorithm used in distributed systems?

  1. QuickSort
  2. Binary Search
  3. Raft
  4. Depth-First Search

Answer: C) Raft

Explanation:

Raft is a consensus algorithm designed to manage replicated logs and coordinate state among nodes. It uses concepts such as leader election, log replication, and safety rules to maintain agreement under its supported failure model.

27. Why are unique request identifiers useful in distributed applications?

  1. They can help detect duplicate requests and support safe retry handling
  2. They guarantee that a server never fails
  3. They eliminate all network latency
  4. They automatically synchronize physical clocks

Answer: A) They can help detect duplicate requests and support safe retry handling

Explanation:

Unique request identifiers allow a service to recognize repeated submissions of the same logical operation. When combined with durable deduplication records or idempotent processing, they can prevent retries from causing unintended duplicate effects.

28. What does idempotency mean in distributed computing?

  1. An operation that must run on every server simultaneously
  2. An operation that cannot return a response
  3. An operation that always changes the system state
  4. An operation that has the same intended effect when repeated with the same input

Answer: D) An operation that has the same intended effect when repeated with the same input

Explanation:

An idempotent operation can be repeated without changing its intended result beyond the effect of the first successful execution. For example, setting a user's status to active can be idempotent, whereas charging a payment repeatedly may not be unless duplicate processing is controlled.

29. What is the purpose of a distributed transaction?

  1. To execute an operation exclusively within a single local variable
  2. To coordinate a logical transaction involving multiple services or resource managers
  3. To prevent data from being stored on disk
  4. To eliminate the need for failure recovery

Answer: B) To coordinate a logical transaction involving multiple services or resource managers

Explanation:

A distributed transaction spans multiple participants, such as databases or services. Transaction protocols or application-level coordination mechanisms help manage completion, consistency, and failure recovery across those participants.

30. What is the purpose of the two-phase commit (2PC) protocol?

  1. To compress data into two files
  2. To synchronize CPU clocks in two steps
  3. To coordinate a transaction's commit or abort decision across participating resource managers
  4. To replicate every file to exactly two servers

Answer: C) To coordinate a transaction's commit or abort decision across participating resource managers

Explanation:

Two-phase commit consists of a prepare phase and a decision phase. Participants first indicate whether they can commit, after which the coordinator communicates the commit or abort decision. Classic 2PC can block if critical participants or the coordinator fail at particular stages.

31. What is a distributed file system?

  1. A system that provides access to files stored across multiple networked machines
  2. A file that can be opened only by one processor
  3. A system that stores all information in CPU registers
  4. A database that cannot communicate over a network

Answer: A) A system that provides access to files stored across multiple networked machines

Explanation:

A distributed file system allows users or applications to access files stored on networked machines. Depending on its design, it may provide replication, fault recovery, access control, caching, and a unified namespace.

32. Which technology is commonly associated with distributed data processing across clusters?

  1. BIOS setup utility
  2. Single-user text editor
  3. Local keyboard driver
  4. Apache Hadoop

Answer: D) Apache Hadoop

Explanation:

Apache Hadoop provides an ecosystem for distributed storage and large-scale data processing. Its core components include HDFS for distributed file storage and MapReduce for distributed batch processing, alongside supporting tools and services.

33. What is the primary function of the MapReduce programming model?

  1. To replace every network protocol with a local function call
  2. To process data through mapping and reduction stages that can run across multiple machines
  3. To prevent parallel processing
  4. To store all application data in processor cache

Answer: B) To process data through mapping and reduction stages that can run across multiple machines

Explanation:

MapReduce divides data processing into map operations that produce intermediate key-value pairs and reduce operations that aggregate values associated with keys. A distributed execution framework schedules tasks and handles aspects of data movement and recovery.

34. What is data partitioning in a distributed database?

  1. Keeping every record only in application memory
  2. Removing all indexes from a database
  3. Dividing a dataset into subsets distributed across different nodes
  4. Converting all records into executable programs

Answer: C) Dividing a dataset into subsets distributed across different nodes

Explanation:

Data partitioning, often called sharding in database systems, distributes portions of a dataset across multiple nodes. It can increase storage capacity and spread workload, but requires appropriate partition keys and mechanisms for routing requests and handling skew.

35. What is a major advantage of using distributed caching?

  1. Reducing repeated access to slower data sources by storing frequently used data in a shared or distributed cache
  2. Guaranteeing that cached data can never become stale
  3. Eliminating the need for data invalidation policies
  4. Preventing all application servers from reading data

Answer: A) Reducing repeated access to slower data sources by storing frequently used data in a shared or distributed cache

Explanation:

Distributed caches can reduce database load and improve response times by serving frequently requested data from memory. Applications must choose suitable expiration, invalidation, and consistency strategies to manage stale or outdated cache entries.

36. What is a heartbeat message commonly used for in distributed systems?

  1. To transfer large database backups exclusively
  2. To synchronize source code versions
  3. To guarantee that a node will never crash
  4. To provide periodic indications that a process or node is responsive

Answer: D) To provide periodic indications that a process or node is responsive

Explanation:

Heartbeat messages help monitoring components assess whether a node appears responsive. Missing heartbeats may indicate a failure, but network delays and congestion can also cause missed messages, so failure detection is generally based on suspicion rather than certainty.

37. What is a circuit breaker pattern in distributed applications?

  1. A physical device required inside every network router
  2. A pattern that temporarily stops calls to a failing dependency and allows recovery checks
  3. A method of disabling all application logging
  4. A technique for making every request wait indefinitely

Answer: B) A pattern that temporarily stops calls to a failing dependency and allows recovery checks

Explanation:

The circuit breaker pattern helps prevent repeated calls to an unhealthy service from worsening failures. It typically uses closed, open, and half-open states to regulate requests and test whether the dependency has recovered.

38. What is service discovery in a distributed system?

  1. A method of deleting inactive users
  2. A process that compiles every microservice into one program
  3. A mechanism that allows clients or services to locate available service instances
  4. A protocol for preventing servers from registering their addresses

Answer: C) A mechanism that allows clients or services to locate available service instances

Explanation:

Service discovery identifies available service instances and their connection details. It is useful when services scale dynamically, move between machines, or are replaced during deployment and recovery.

39. What is the purpose of a distributed lock?

  1. To coordinate access to a shared resource among processes running on different nodes
  2. To prevent every machine from accessing the network
  3. To ensure that no process can execute concurrently anywhere in the system
  4. To permanently reserve all CPU resources for one server

Answer: A) To coordinate access to a shared resource among processes running on different nodes

Explanation:

A distributed lock helps coordinate access to shared resources when multiple processes may attempt conflicting operations. Correct implementations must consider lease expiration, failures, and stale lock holders; fencing tokens can help protect resources from delayed operations by former lock holders.

40. What is the main goal of distributed tracing?

  1. To replace all application testing
  2. To guarantee that network requests have zero latency
  3. To synchronize every physical clock exactly
  4. To follow a request across multiple services and identify where time is spent or failures occur

Answer: D) To follow a request across multiple services and identify where time is spent or failures occur

Explanation:

Distributed tracing records the path of a request through multiple components using traces and spans. It helps engineers investigate latency, service dependencies, and failures in applications composed of many distributed services.

41. What is the purpose of a quorum in a replicated distributed system?

  1. To guarantee that all network messages arrive simultaneously
  2. To require agreement or participation from a specified number of replicas for an operation
  3. To disable replication permanently
  4. To ensure that every node has the same IP address

Answer: B) To require agreement or participation from a specified number of replicas for an operation

Explanation:

A quorum is a required subset of participating replicas used to make decisions or complete operations. For example, a system with five replicas may require three acknowledgments. The guarantees depend on the read and write quorum rules, replication protocol, and failure model.

42. What is a split-brain condition in a distributed system?

  1. A process in which data is compressed into two formats
  2. A situation in which all replicas are permanently deleted
  3. A condition in which separated parts of a cluster each believe they are the active authority
  4. A method of increasing the number of CPU cores

Answer: C) A condition in which separated parts of a cluster each believe they are the active authority

Explanation:

Split brain can occur when a cluster is partitioned and multiple groups independently assume leadership or accept conflicting writes. Quorum-based decisions, fencing mechanisms, and carefully designed leader-election protocols help reduce the risk of conflicting operations.

43. What is the purpose of a retry mechanism in distributed communication?

  1. To attempt an operation again after a potentially temporary failure
  2. To guarantee that every failed operation succeeds on the next attempt
  3. To eliminate the need for request timeouts
  4. To force all requests to use the same server

Answer: A) To attempt an operation again after a potentially temporary failure

Explanation:

Retries can recover from transient failures such as temporary connection problems. They should use appropriate limits, timeouts, and often exponential backoff with jitter. Idempotency or duplicate-request protection is important when a timed-out operation may already have succeeded.

44. What is the purpose of exponential backoff in distributed systems?

  1. To increase the number of retries without any limit
  2. To send every retry immediately
  3. To disable communication after the first failure permanently
  4. To increase the delay between successive retries, subject to configured limits

Answer: D) To increase the delay between successive retries, subject to configured limits

Explanation:

Exponential backoff increases the waiting period between repeated attempts. This reduces pressure on an overloaded or recovering service. A maximum delay, retry limit, and random jitter are commonly used to prevent synchronized retry bursts.

45. Which security mechanism helps protect communication between distributed services from interception and tampering?

  1. Unencrypted communication over public networks
  2. TLS (Transport Layer Security)
  3. Disabling authentication on every endpoint
  4. Sharing the same password across all users

Answer: B) TLS (Transport Layer Security)

Explanation:

TLS protects communication confidentiality and integrity and can authenticate the server. Mutual TLS can also authenticate clients. Secure distributed systems additionally require authorization, credential management, and appropriate identity verification.

46. What is the primary purpose of access control in a distributed system?

  1. To allow every user to access every resource
  2. To eliminate the need for user identities
  3. To restrict resource and operation access according to defined permissions
  4. To guarantee that servers never experience downtime

Answer: C) To restrict resource and operation access according to defined permissions

Explanation:

Access control determines which users, services, or processes may access resources and perform particular actions. Authentication establishes identity, while authorization evaluates whether that identity has permission to perform a requested operation.

47. What is a microservices architecture?

  1. An architecture in which an application is divided into independently deployable services communicating through defined interfaces
  2. A design that requires all application functionality to reside in one executable process
  3. A system that prohibits network communication between components
  4. A database storage format that cannot be updated

Answer: A) An architecture in which an application is divided into independently deployable services communicating through defined interfaces

Explanation:

Microservices organize an application as a collection of services, each responsible for a defined capability. Services can often be deployed and scaled independently, but this architecture introduces challenges involving communication, observability, data consistency, and operational coordination.

48. What is the primary purpose of a distributed scheduler?

  1. To ensure that every task executes on the same node
  2. To disable parallel execution across a cluster
  3. To synchronize physical clocks without a network
  4. To assign tasks to available computing resources across multiple nodes

Answer: D) To assign tasks to available computing resources across multiple nodes

Explanation:

A distributed scheduler assigns tasks across a set of computing resources based on factors such as availability, capacity, dependencies, and scheduling policies. It may also handle retries, placement constraints, and workload balancing.

49. Which statement best describes the difference between parallel computing and distributed computing?

  1. Parallel computing always requires geographically distant machines
  2. Parallel computing focuses on simultaneous task execution, often using closely coordinated processors, while distributed computing involves independent networked computing entities coordinating through communication
  3. Distributed computing cannot execute tasks simultaneously
  4. Both terms always refer to exactly the same architecture

Answer: B) Parallel computing focuses on simultaneous task execution, often using closely coordinated processors, while distributed computing involves independent networked computing entities coordinating through communication

Explanation:

Parallel computing emphasizes executing multiple operations at the same time, often within one machine or a tightly coupled system. Distributed computing emphasizes coordination among independent networked computers. The two concepts overlap because distributed systems can perform parallel computations.

50. A distributed application stores data on five replicas. Its write operation requires acknowledgments from at least three replicas. One replica is unavailable, and the remaining four are reachable. Assuming no other failures and that the system permits writes with the required quorum, what is the minimum number of successful replica acknowledgments needed for the write to proceed?

  1. One acknowledgment
  2. Two acknowledgments
  3. Three acknowledgments
  4. Five acknowledgments

Answer: C) Three acknowledgments

Explanation:

The configured write quorum requires at least three successful acknowledgments. Because four replicas remain reachable, the system can satisfy that requirement if three of them successfully acknowledge the write. The unavailable replica does not automatically prevent the operation. However, quorum size alone does not establish every consistency guarantee; those also depend on the replication protocol and read/write rules.

Comments and Discussions!

Load comments ↻



Top MCQs
Top Programs/Examples
Top Tutorials
About
Student's Section
Subscribe

Copyright © 2026 www.includehelp.com. All rights reserved.