Distributed System KCS-077 Aktu Btech Short Question, Quantum Book Pdf

Learn about the B.Tech AKTU Quantum Book Short Question Notes on Distributed System. Learn the fundamentals of decentralised computing and networking, as well as how to develop robust and scalable distributed applications.

Dudes 🤔.. You want more useful details regarding this subject. Please keep in mind this as well.

Important Questions For Distributed System:
*Quantum                *B.tech-Syllabus
*Circulars                 *B.tech AKTU RESULT
* Btech 4th Year     * Aktu Solved Question Paper

Unit-I: Characterization of Distributed System (Short Question)

Q1. Discuss the role of file system in distributed system. 

Ans. Following are the role of file system in distributed system: 

• 1. Managing files and folders on various machines.  
• 2. To transfer information and files among network users in a controlled and authorized manner. 

Q2. List the goals of distributed systems. 

Ans. Goals of distributed system are: 

• i. Improved system performance 
• ii. Resource sharing
• iii. Improved reliability and availability 
• iv. Modular expendability 

Q3. List out the main challenges of distributed systems. 

Ans. Following are the main challenges of distributed system: 

• 1. Security 
• 2. Scalability  
• 3. Failure handling
• 4. Concurrency
• 5. Transparency

Q4. What are logical clocks ? Why does a logical clock need to be implemented in distributed systems?

Ans. In a distributed system, a logical clock is a technique for capturing temporal and causal linkages. Because distributed systems lack a physically synchronous global clock, a logical clock enables global ordering of events from many processes.

Q5. List out some issues in distributed file systems. 

Ans. Issues in distributed file system: 

• 1. Naming and name resolution 
• 2. Cache consistency 
• 3. Availability  
• 4. Scalability 
• 5. Semantics

Q6. What are the web challenges involved in distributed system?  

Ans. Following are the web challenges involved in distributed system: 

• 1. Naming  
• 2. Access control
• 3. Security
• 4. Performance management 
• 5. Availability 

Q7. Explain system model. 

Ans. In a single descriptive model, system models express common attributes and design choices for distributed systems. System models are of two types: 

• i. Architectural models 
• ii. Fundamental models 

Q8. Why would you design a system as a distributed system? List some advantages of distributed systems. 

Ans. 1. A distributed system is a collection of separate computers that collaborate and operate concurrently. This permits the entire system to continue to function even if one or more computers fail. 

2. The distributed system is scalable, which means that it can be easily modified to handle changes in the number of users and resources. As a result, a distributed system can be much larger and more powerful than the sum of separate systems. 

3. For the reasons stated above, we would build a system as a distributed system. 

Advantages of distributed systems:

• i. Shareability
• ii. Expandability 
• iii. Improved performance  
• iv. Improved reliability
• v. Flexibility 

Q9. Name the algorithm for causal ordering of messages. 

Ans. Algorithms for causal ordering of messages are: 

• i. Birman-Schiper-Stephenson algorithm
• ii. Schiper-Eggli-Sandoz algorithm

Q10. What is causal ordering of message ? 

Ans. The casual ordering of messages is concerned with the concept of keeping the same casual relationship that exists between the “message send” and “message receive” events. 

Q11. Define global state. 

Ans. A distributed computation’s global state is the set of local states of all individual processes involved in the computation as well as the state of the communication channels. The global state of the system is a collection of a processing system’s local states. 

GS = {LS₁, LS₂, LS₃, ….. LS_N}

where N is the number of sites in the system. 

Q12. List three properties of distributed systems. 

Ans. Properties of distributed system: 

• 1. Resource sharing
• 2. Scalability 
• 3. Fault tolerance

Q13. What is the advantage if your server side processing uses threads instead of a single process ?

Ans. Advantages: 

• 1. Threads are created and destroyed much faster than process. 
• 2. It is faster to switch between threads than to switch between processes.
• 3. Threads share data easily. 

Q14. What is a proxy ? Give an example of where a proxy can be used. 

Ans. A proxy (or proxy server) is a server that serves as a conduit between a user’s computer and the internet. A proxy server checks incoming client requests and sends them to other servers for further communication. A web browser or network can employ proxy servers to improve privacy. For example, queries for a website sent over a proxy may aid in hiding the client’s IP address from the web server. 

Q15. What is the purpose of a firewall ? 

Ans. Purpose of firewall: 

• 1. To protect computers and private networks from internet threats. 
• 2. Packet filtering. 
• 3. To record all the information from packets that passes through it. 

Q16. What were the reasons that middleware moved from distributed objects to distributed components ? 

Ans. Following are the reasons that middleware moved from distributed objects to distributed components: 

• 1. Implicit dependencies: The dependencies the object has on other objects in the distributed configuration. 
• 2. Interaction with the middleware: A clear separation is required between code related to middleware operation and code relevant to the application. 
• 3. Lack of separation of distribution concerns: The complexity of working with distributed system services should be kept from the programmer as much as feasible. 
• 4. No support for deployment: Middleware platforms should have deployment capability by default, allowing distributed software to be installed in the same way that software for a single computer is.

Q17. Name two mechanisms that can be used to ensure performance in distributed systems.

Ans. Two mechanisms that can be used to ensure performance in distributed systems are: 

• 1. Fault tolerance 
• 2. Availability 

Q18. What are the differences between a URL, URI and URN?

Ans. 

Q19. What is consistent cut and inconsistent cut?

Ans. Consistent cut: A cut is called consistent, if for each event that it contains, it also includes all events all events causally ordered before it. Let a, b be two events in a distributed system. Then 

(a ∊ consistent cut C) ∧ (B < a) ⇒ b ∊ C 

Thus, for a message m, if the state following receive (m) belongs to a consistent cut, then the state following send (m) also must belong to cut. 

Inconsistent cut: A cut C is inconsistent iff any of the following three basic conditions is true:

1. e is sending event of a message and e’ is the receipt event of the same message ∧ e’ ∊C ∧ e ∉ C.

Q20. What is termination detection problem ?

Ans. The termination detection challenge entails determining if an ongoing distributed computation has completed all of its tasks. When a distributed computation terminates implicitly, that is, once the calculation has completed all of its activities, no one process is aware of the termination, the termination detection problem occurs. As a result, a different technique must be used to detect computation termination.  

Q21. Explain desirable features of a good message passing system. 

Ans. Following are the desirable features: 

• i. Simplicity
• ii. Uniform semantics 
• iii. Efficiency 
• iv. Atomicity 
• v. Ordered delivery
• vi. Reliability 
• vii. Flexibility 
• viii. Security 

Unit-II: Distributed Mutual Exclusion (Short Question)

Q1. What do you mean by mutual exclusion in distributed system ? What are the requirements of a good mutual exclusion algorithm ?

Ans. Mutual exclusion: 

• 1. Mutual exclusion is a problem that occurs when a process is dependent on a shared resource that can only be used by one process at a time. 
• 2. Access to shared resources is not permitted at the same time. 
• 3. The mutual exclusion algorithm ensures that only one request at a time can access the crucial section (CS). 

Requirements of good mutual exclusion algorithm: 

• 1. Freedom from deadlocks: Two or more sites should not endlessly wait for messages that will never arrive. 
• 2. Freedom from starvation: A site should not be forced to wait endlessly to run CS; rather, each asking site should be given a limited time to conduct CS. 
• 3. Fairness: Fairness dictates that requests must be executed in the order in which they arrive in the system. 
• 4. Fault tolerance: A fault-tolerant mutual exclusion algorithm may recognise itself after a failure and continue to work without any (prolonged) disturbances. 

Q2. Define deadlock detection in distributed systems. 

Ans. The system does not seek to prevent deadlock in deadlock detection, but instead allows processes to request resources and wait for each other in an uncontrolled way.

Q3. What is distributed deadlock ?

Ans. A deadlock is a circumstance in which a collection of processes is stalled because each process is holding a resource and waiting for another process to obtain another resource. Deadlock detection in a distributed DBMS is more difficult because it includes multiple sites. In order for the entire system to detect a deadlock situation in a distributed DBMS, a global wait-for graph (GWFG) must be drawn. 

Q4. Explain token based algorithm.

Ans. 1. A unique token is shared by all sites in the token-based algorithm. If a site has the token, it is permitted to enter its CS (Critical Section). 

2. Sequence numbers are used instead of timestamps in token-based algorithms. 

3. Each token request has a sequence number, and each site’s sequence number progresses independently. Every time a site requests a token, it increments its sequence number counter. 

Q5. What are the necessary conditions for deadlock ?  

Ans. Necessary conditions for deadlock: 

• i. Mutual exclusion 
• ii. Hold and wait 
• iii. No preemption 
• iv. Circular wait 

Q6. Define WFG.

Ans. WFG stands for wait-for graph. The state of a distributed system can be represented by a directed graph known as a wait-for graph (WFG). A system is blocked if and only if the WFG contains a directed cycle or knot. 

Q7. What are the different classes of distributed deadlock detection algorithms ?

Ans. Following are different classes of distributed deadlock detection algorithms: 

• i. Path-pushing algorithm 
• ii. Edge-chasing algorithm 
• iii. Diffusion computation 
• iv. Global state detection

Q8. Explain edge chasing algorithm. 

Ans. To identify the cycle in edge chasing algorithms, special messages called probes are transmitted around the edges of the wait-for graph. When a blocked process gets the probe, the probe is propagated down the outgoing edges of the wait-for graph. 

Q9. Write a short note on hierarchical deadlock detection algorithm. 

Ans. In this, sites are logically arranged in hierarchical fashion and a site is responsible for detecting deadlocks involving only its children sites. It is of two types: 

• i. The Menasce-Muntz algorithm 
• ii. The Ho-Ramamoorthy algorithm

Q10. What are the types of token based algorithm ?

Ans. Types of token based algorithm are: 

• i Suzuki-Kasami’s broadcast algorithm. 
• ii. Singhal’s algorithm 

Q11. What are the types of non-token based algorithm ?

Ans. Types of non-token based algorithms are: 

• i. Lamport’s algorithm Distributed Mutual Exclusion 
• ii. Ricart-Agarwala algorithm 
• iii. Maekawa’s algorithm 

Q12. What are the performance matrices used to measure performance of mutual exclusion algorithm?

Ans. Performance matrices to measure performance of mutual exclusion algorithm are: 

• i. Response time 
• ii. Synchronization delay 
• iii. Number of messages per critical section 
• iv. System throughput 

Q13. Write are the design issues of Suzuki-Kasami algorithm ? 

Ans. Design issues of Suzuki-Kasami algorithm are: 

• 1. It differentiates between outdated and current request messages. 
• 2. It decides which site has a request for the critical section that is out of date. 

Unit-III: Agreement Protocols (Short Question)

Q1. What are distributed shared memory design issues?

Ans. Design issues in distributed shared memory are: 

• i. Granularity 
• ii. Structure of shared memory space
• iii. Memory coherence
• iv. Data location and access 
• v. Block replacement policy 
• vi. Thrashing 
• vii. Heterogeneity

Q2. Why clocks need to be synchronized ?

Ans. There is no global clock or shared memory in distributed systems. Each processor has its own internal clock, which can easily diverge by several seconds each day, accumulating large inaccuracies over time. This poses a major difficulty for applications that rely on clock synchronization. Clocks must be synchronized to correct clock drift. 

Q3. What do you mean by agreement protocol ?

Ans.

• 1. The agreement protocol refers to the process of transmitting and reaching an agreement to all locations. 
• 2. Agreement protocols are extremely important in distributed systems for error-free communication among diverse sites. 
• 3. Faulty processors are more likely in distributed systems. A defective processor may result in incorrect message communication, no response to a message, and so on. 
• 4. The agreement protocols enable non-defective processors in the distributed system to achieve a common agreement, regardless of whether other processors are faulty or not. 
• 5. The agreement protocol is used to get a shared agreement among the processors. 

Q4. State Byzantine agreement problem. 

Ans.

• 1. In the Byzantine agreement problem, a randomly selected processor, known as the source processor, broadcasts its initial value to all other processors. 
• 2. A solution to the Byzantine agreement problem should meet the following objectives:
  • i. Agreement: All non-faulty processors agree on the same value. 
  • ii. Validity: If the source processor is not faulty, then the value agreed upon by all non-faulty processors should be the source’s initial value. 
  • iii. Termination: Each non-faulty processor must eventually decide on a value.

Q5. Explain the concept of shared memory.

Ans. 1. Distributed shared memory (DSM) is a type of memory architecture in which several (physically separate) memories can be addressed as if they were in the same (logically shared) address space. 

2. The shared memory paradigm creates a virtual address space that is shared by all nodes. 

3. DSM is basically a tool for any distributed application that allows direct access to specific shared data items. 

Q6. What are the various aspects for recognizing the agreement protocol ?

Ans. Various aspects for recognizing the agreement protocols are:

• i. Synchronous and Asynchronous computations  
• ii. Model of processor failures  
• iii. Authenticated and non-authenticated messages
• iv. Performance aspects 

Q7. Give the two conditions that a clock synchronization algorithm should satisfy. 

Ans. Following are the two conditions that a clock synchronization algorithm should satisfy:

• i. The clock values of all non-faulty processes must be approximately equal at all times. 
• ii. The amount by which the clock of a non-faulty process is modified during each resynchronization is limited. 

Q8. What are the services offered by a distributed file system ? 

Ans. Services offered by a distributed file system are: 

• i. Storage service 
• ii. Flat-file service 
• iii. Name service

Q9. What are the requirements of a distributed file system ?  

Ans. Requirements of a distributed file system are: 

• i. Transparency 
• ii. Concurrent file updates 
• iii. File replication 
• iv. Hardware and operating system heterogeneity 
• v. Fault tolerance 
• vi. Security 
• vii. Efficiency

Q10. List out the mechanisms for building Distributed File System (DSF). 

Ans. Mechanisms for building DFS: 

• i. Mounting 
• ii. Caching 
• iii. Hints 
• iv. Bulk data transfer 
• v. Encryption 

Q11. What are the algorithms for implementing distributed shared memory ? 

Ans. There are four basic algorithms to implement DSM systems: 

• i. The central-server algorithm 
• ii. The migration algorithm 
• iii. The read-replication algorithm 
• iv. The full-replication algorithm

Unit-IV: Failure Recovery in Distributed System (Short Question)

Q1. Compare and contrast static and dynamic vote protocol. 

Ans. 

Q2. Define fault and failure. What are different approaches to fault tolerance ?  

                                                                  OR

Define fault and failure in distributed system. 

Ans. Faults: A flaw is a physically abnormal condition. Faults are caused by design faults, manufacturing issues, damage fatigue or other deterioration, and external disturbances. 

Failure: A system fails when it fails to perform its functions in the manner specified. An erroneous system state is one that could result in a system failure through a sequence of valid state transactions.  

Different fault tolerance approaches: 

• 1. Replication: 
  • a. Replication refers to the process of producing and maintaining numerous copies of data items or processes across multiple nodes. 
  • b. As a result, if one node fails, data from the other node will be accessible to the user. 
  • c. High data availability and performance are provided by replication. 
• 2. Checkpointing: 
  • a. Checkpointing can be used to achieve fault tolerance. 
  • b. Checkpointing is the process of saving the system’s consistent state on a dependable storage medium on a regular basis. A checkpoint is an occurrence when a system is in a consistent state. 
  • c. Checkpointing is generally used to avoid losing all useful processing done before to the occurrence of a failure. 
  • d. In the event of a fault, checkpoint allows programme execution to be continued from a previous consistent state rather than from the beginning. 

Q3. What do you mean by commit protocol ?

Ans. 1. Commit protocols in distributed systems provide atomicity across sites, i.e., when a transaction executes at several sites, it must either be committed at all sites or aborted at all sites.  

2. The purpose of commit protocols is for all concerned parties to agree to either commit or abort a transaction. 

Q4. What is error ?  

Ans. An error is a component of the system state that deviates from its intended value. An error is a symptom of a flaw in a system that could lead to system failure. 

Q5. Give classification of failures. 

Ans. Failures can be classified as: 

• i. Process failure 
• ii. System failure 
• iii. Secondary storage failure 
• iv. Communication medium failure  

Q6. Why one phase atomic commit protocol is inadequate ? 

Ans. A one-phase atomic commit protocol is insufficient because it does not allow a server to make a unilateral choice on a transaction when the client asks one.  

Q7. What are the properties that influence ability of distributed file system to tolerate faults ? 

Ans. Properties are:

• i. Recoverability 
• ii. Availability 
• iii. Robustness 

Q8. What are the approaches to implement resilient process ?

Ans. Approaches to implement resilient process: 

• i. Backup processes 
• ii. Replicated execution

Q9. Mention approaches to implement backward error recovery. 

Ans. Approaches to implement backward recovery:

• i. Operation based approach 
• ii. State based approach

Q10. Differentiate between local and global checkpointing. 

Ans. 

Q11. What is pessimistic logging ?

Ans. In pessimistic logging, an incoming message is logged before it is processed.  

Q12. What is optimistic logging?

Ans. In optimistic logging, processors continue to compute while the messages received are stored in volatile storage and logged at regular intervals.

Unit-V: Transaction and Concurrency Control (Short Question)

Q1. What is atomic commit protocol ?

Ans. The atomic commit protocol (ACP) is a technique used by database managers to ensure that all sub-transactions are consistently committed or cancelled. Each server applies local concurrency control to its own object, ensuring that transactions are serialized both locally and globally.  

Q2. How shadow versions are helpful in recovery ?

Ans. The shadow versions technique is another method for organizing a recovery file. It employs a map to locate versions of the server’s objects stored in a file known as a version store. When a server is replaced after a crash, its recovery manager reads the map and utilizes the map’s metadata to locate the objects in the version store. 

Q3. What are the differences between a local call and a remote call ? 

Ans. 

Q4. State timestamp ordering. 

Ans. 1. The timestamp ordering technique is used to place transactions in chronological order based on their timestamps. The transaction order is just the ascending sequence of transaction creation. 

2. The older transaction has a greater priority and gets completed first. The protocol employs system time or a logical counter to determine the timestamp of the transaction. 

3. The timestamp ordering protocol also keeps track of the last’read’ and ‘write’ operation on a piece of data. 

Q5. Explain the effect of replicated data in transactions. 

Ans. Effect of replicated data in transaction: 

• 1. It enhances the availability and performance of data. 
• 2. It makes recovery more complicated. 

Q6. What are the different validation conditions for optimistic concurrency control ?

Ans. Different validation conditions for optimistic concurrency control: Let T_i and T_j be the two transactions. For a transaction T_j to be serializable with respect to an overlapping transaction T_i, their operations must confirm to the following rules/conditions:

• 1. T_i must not read objects written by T_j. 
• 2. T_j must not read objects written by T_i. 
• 3. T_i must not write objects written by T_j and T_j must not write objects written by T_i. 

Q7. What do you mean by transactions ?

Ans. A transaction is a sequence read, write or compute operation on any data object of a database. 

Q8. What are the properties of transactions ?

Ans. Properties of transaction: 

• i. Atomicity
• ii. Consistency 
• iii. Isolation
• iv. Durability 

Q9. What are the advantages of nested transaction ?

Ans. Advantages of nested transaction: 

• i. Subtransactions at one level of the hierarchy may operate concurrently with other subtransactions at the same level. This allows for more concurrency in a transaction. 
• ii. Subtransactions can commit or abort on their own. A group of nested subtransactions is possibly more robust than a single transaction. 

Q10. What do you mean by two phase locking ?

Ans. A two phase locking technique is a dynamic locking scheme in which a transaction seeks a lock on a data item when it requires it. It has two phases: 

• i. Growing phase 
• ii. Shrinking phase  

Q11. Write down the main categories of concurrency control. 

Ans. Categories of concurrency control are: 

• i. Optimistic  
• ii. Pessimistic  
• iii. Semi-optimistic 

Q12. What are main task of the group membership service ?

Ans. A group membership service has four main tasks which are as follows:

• i. Providing an interface for group membership changes. 
• ii. Implementing a failure detector. 
• iii. Notifying members of group membership changes. 
• iv. Performing group address expansion. 

Q13. Why two phase locking suffers from cascaded roll back ?

Ans. When a transaction may be rolled back after releasing the locks on some data objects and other transactions have read those updated data objects, two phase locking suffers from the cascaded roll back problem. 

Q14. What is static locking ?  

Ans. Before performing any action on the data objects, a transaction acquires locks on all of them using static locking.  

Q15. When a transaction is called well-formed ?  

Ans. A transaction is well-formed if it: 

• i. Locks a data object before accessing it. 
• ii. Does not lock a data object more than once. 
• iii. Unlocks all the locked data objects before it completes. 

Q16. What are the advantages of distributed database system ?

Ans. Advantages of distributed database system: 

• i. Sharing 
• ii. Higher system availability (reliability) 
• iii. Improved performance 

Q17. Where distributed transactions can be used ?

Ans. Web services-based systems implement distributed transactions. Typically, these transactions employ compensating transaction principles, optimism, and isolation without locking. 

Distributed System Btech Quantum PDF, Syllabus, Important Questions

Distributed System Quantum PDF | AKTU Quantum PDF:

AKTU Important Links | Btech Syllabus

Important Links-Btech (AKTU)