Survey Maker Survey Templates Contact Privacy & Security

Ds exam

Which of the following statements represent false assumptions about a Distributed System?

Latency is zero

Network is reliable

Uses the asynchronous communication model

Topology does not change

Which of the following statements are true about time-space diagrams?

They show the time evolution of a distributed system showing the processes, atomic events execution and the messages exchanged by processes

Together with the concept of “cut” they define the Lamport scalar time

They show the time usage of the memory space in the DS

They help identifying security problems in a DS

Which of the following objectives should be achieved by concurrency control Algorithms?

Guarantee that multiple transactions can be executed simultaneously while still being isolated

Schedule individual read/write operations as to achieve transaction isolation and consistency

Increase data security and privacy across several transaction

The global state of a DS (choose the right answer):

Is represented by the state of all processes and of all communication channels

Is represented by the state of all process and their sent messages

Is represented by a snapshot of the memory of the distributed systems’s processing nodes

Is represented by a snapshot of the messages in transit in the communication channels

Which of the following statements referring to Message Queues communication between processes are correct?

It is a type of direct communication

It is a type of indirect communication

The communication entities must be active at the same time

The communication entities may not be active at the same time

In the transaction related pessimistic concurrency control algorithm (choose the right answer):

Transactions are executed without checking for serializability condition

Transactions operation are synchronized before they are executed

Transactions operation are synchronized after they are executed

Transactions are executed and the check for serializability is done right after committing

The Mean Time To Failure (MTTF) is a measure that is specific to:

Availability (apare in curs)

Performance

Reliability

Consistency

In a Distributed Database Environment, the Global Conceptual Schema contains information about:

The locations where the individual tables are stored

The database tables, their columns, data types, primary and secondary keys, etc.

How the databases are fragmented and where the fragments are located

Which of the following is true about the Chandy-Lamport snapshot algorithm for FIFO channels (Course8)?

The marker is sent by a site before taking its snapshot

The marker is sent by a site after taking its snapshot

Which of the following are true for a P2P system?

A P2P system consists of many server processes and many client processes

A P2P system consists of server process and many client processes

All P2P processes offer the same interfaces

All P2P processes run on the same program

Which of the following statements describe the vertical fragmentation of a database?

Splits table and groups the columns into fragments

Splits table and groups the rows based on the values of one or more columns

Splits tables and groups the rows based on the rows of another table

Splits tables and groups the rows based on the columns of another table

To improve low utilization ratio of server machines in virtual clusters it is recommended to:

Increase servers memory

Perform server consolidation

Allocate the execution of the same task on more servers

Add new server machines

Which type of cloud service provides resources to allow the user to develop, deploy, and manage the execution of their applications?

Infrastructure as a Service (IaaS)

Platform as a Service (PaaS)

Software as a Service (SaaS)

CAP theorem states that every distributed system can provide only two of the following properties:

Consistency, Availability, Performance

Consistency, Availability, Partition-Resilience

Consistency, Access, Performance

Concurrency, Availability, Partition-Resilience

Which of the following are true about logical clocks

They capture the event casualty in a more accurate way than physical time

They measure the time between two distributed system events

The time measured is expressed in nanoseconds

They can be used to timestamp the events of the distributed system execution

In RPC mode a stub is:

A remote procedure implemented by a server

A proxy of a remote procedure

A server side dispatcher module used to identify the procedure to be executed

Which of the following statements are true in case of full virtualization:

Guest OS knows that is running in the virtualized environment

Guest OS is completely abstracted from the underlying infrastructure

Guest OS is used without any modification

Guest OS is modified to replace non-virtualized instructions with hypercalls

Bi-directional streaming is

A feature of all RPC technologies

A feature of gRPC

A feature of virtualization

Which of the following is a feature of a DS?

If one node fails, the rest of the DS nodes will stop execution

If one node fails, only the directly connected nodes to the failed node will stop execution

If one node fails, the rest of the nodes will continue to operate

Which of the following represent benefits of resource virtualization?

Low energy consumption

Elastic usage of hardware resource

Server consolidation

Which of the following are true for a Distributed Transaction (DT)?

A DT must be of all or nothing semantics

A DT generates a transition of a database from a consistent state to a consistent state

DT’s effect is transparent to other transactions until it commits

A deadlock situation may occur in all concurrent executed DTs no matter of the concurrency control technique

Which of the following is not true about Java RMI?

Parameters can be passed both by reference and by value

The distributed object interface must be deployed both on client and server

The distributed garbage collector counts both local and remote references to an object

RMI Registry is reference implementation for a naming service

Which of the following must be satisfied by the global state of a distributed system to be consistent?

Every message recorded as sent by the sender processor should be in the communication channel or already received by the destination process

If a message is not recorder as sent it cannot be in the communication channel or already received

There should be no messages in the communication channels

The order of the sent messages recorded at the sender processes should be the same at the receiving processes

A database is said to be distributed if:

All its stored data is localized at a single site

Any of its tables are stored at different sites

One or more of its tables are replicated and their copies are stored at different sites

One or more of its tables are fragmented and the fragments are stored at different sites

In a two-phase locking algorithm:

In the Growing Phase, the scheduler acquires all the locks a transaction is needed

In the Growing Phase, the scheduler acquires all the locks needed for a write operations

In the Shrinking Phase the scheduler releases the locks as each operation completes

In the Shrinking Phase the scheduler releases all locks at once after transaction terminates

Which type of failures are also known as Byzantine failures?

The omission failures

The arbitrary failures

The timing failures

The masking failures

An explicit wait-for graph may be used for implementing (choose the right answer):

A deadlock detection technique for lock-based concurrency control

A deadlock detection technique for optimistic concurrency control

In the One Phase Atomic Control Protocol:

The coordinator takes the commit or abort decision

The participants vote commit or abort

The coordinator together with the participants take the commit or abort decision using voting

The participants cannot decide to abort the transaction on their own

System Throughput is:

An indicator of availability

An indicator of the number of queries/ requests per second

And indicator of durability

In transaction implementation technique based on write-ahead logs (check all true statements):

Each of the participant nodes keeps its own log of transaction changes

Only the coordinator keeps a log of transaction changes

The data is effectively changed only after the log is written successfully

The data is changed and then, the modification is written on the log

Care din urmatoarele afirmatii sunt adevarate despre protocolul Two-Phase Commit Protocol (2PC) pentru tranzactii distribuite?

Fiecare participant poate sa decida sa faca commit sau abort la cererea coordonatorului si sa execute partea lui de tranzactie ca urmare a deciziei luate

Fiecare participant poate sa decida sa faca commit sau abort la cererea coordonatorului si sa execute partea lui de tranzactie numai dupa ce a primit si decizia coordonatorului

Dupa ce a votat pentru Commit un participant poate doar sa astepte dupa decizia coordonatorului fara a avea alte interactiuni cu acesta

Erorile sunt mai usor de detectat in sisteme distribuite asincrone decat in sistemele distribuite sincrone

Adevarat

Fals

Care din urmatoarele sunt sarcini ale Hypervisor-ului in cazul unui sistem virtualizat?

Virtualizeaza resursele hardware fizice in resurse virtuale

Administreaza resursele hardware a unei masini fizice

Izoleaza fiecare masina fizica de celelalte din sistem

Intr-un sistem distribuit, replicarea datelor imbunatateste (selecteaza raspunsul corect):

Consistenta datelor

Toleranta la erori

Atat consistenta datelor cat si toleranta la erori

Care din urmatoarele afirmatii sunt adevarate despre tranzactiile distribuite nested?

O tranzactie distribuita nested poate avea sub-tranzactii copii (children sub-transaction)

Sub-tranzactiile copii nu pot sa aiba la randul lor alte sub-tranzactii

Sub-tranzactiile copii se pot executa in paralel pe masini diferite

Pentru motive de consistenta, sub-tranzactiile copii trebuie sa fie executate una dupa alta

Cand o tranzactie copil face commit rezultatele ei sunt vizibile pentru tranzactia parinte

Care dintre urmatoarele afirmatii sunt adevarate despre tranzactiile distribuite flat?

Operatiile tranzactiilor pot fi executate pe servere diferite

Operatiile tranzactiilor nu se executa in paralel

Cateva din operatiile tranzactiilor, dar nu toate, pot fi executate in paralel

Fiecare operatie a tranzactiei trebuie sa fie executata inainte de invocarea unei alte operatii

Decuplarea in timp si spatiu sunt proprietati a:

Comunicarii indirecte

Comunicarii directe

Apelul la distanta a unei metode

Comunicarii request-reply

Care din urmatoarele afirmatii sunt adevarate despre Transactions Concurrency Control?

Permite unui set de tranzactii sa fie executate in mod simultan pe resurse partajate (shared resources)

Este aplicabil doar pentru tranzactii distribuite

Impune consistenta intre resursele implicate in tranzactie

Cresterea numarului de tranzactii pe secunda intr-un sistem distribuit este un exemplu de (selecteaza raspunsul corect):

Up-scaling a sistemului

Out-scaling a sistemului

Google Docs este un tip de:

Infrastructura ca

Serviciu Software ca Serviciu

Platforma ca Serviciu

Niciunul din raspunsurile oferite

Care dintre proprietatile ACID descrie faptul ca multiple tranzactii se pot executa independent fara nici o interferenta?

Atomicitate

Consistenta

Izolare

Durabiliate

Care din urmatoarele sunt proprietati a tehnicii de ordonare bazata pe timestamp-uri in cazul controlului concurentei?

Nu sunt operatii conflictuale in cazul acestei tehnici

Ordonarea bazata pe timestamp-uri este o tehnica optimista de control a concurentei

Fiecarei tranzactii distribuite ii este asignat un timestamp unic global

Niciuna dintre raspunsuri

In cazul Remote Procedure Call (RPC) parametrii de apel sunt trimisi (bifeaza raspunsul corect):

Prin valoare

Prin referinta

In ambele moduri

Verificarea conflictului intre operatii in cazul algormitmului de Two Phase Locking (2PL) pentru controlul concurentei este sarcina carei componente:

Transaction Manager

Scheduler

Data Manager

Relocarea unei componente a unui sistem distribuit fara a instiinta clientul este numita:

Transparenta de acces

Transparenta de migrare

Transparenta de locatie

Transparenta de concurenta

In cazul Remote Procedure Call reprezentantul pe partea de client a unei proceduri remote este numit:

Skeleton

Stub

Interfata remote

Interfata client

Abilitatea unui sistem distribuit de a opera si a livra servicii chiar si in prezenta erorilor este numita:

Availability

Reliability

Consistency

Efficiency

Care din urmatoarele sunt adevarate despre fragnemtarea tabelelor in design-ul unei baze de date distribuita:

Fragmetele tabelelor trebuie sa fie stocate in site-uri diferite

Fragmetele tabelelor nu trebuie sa fie replicate

Fragmentarea tabelelor ajuta in cazul recuperarii din dezastre (disaster recovery)

Rolul unei functii de timp logic intr-un sistem ceasuri logice (system of logical clocks) este sa:

Asigneze instante de timp evenimentelor

Updateaza timpul logic local pentru un proces

Updateaza timpul logic global pentru un proces

Performanta sistemelor virtualizate din punct de vedere a timpului de executie este mai buna decat performanta sistemelor ne virtualizate

Adevarat

Fals

Un patern arhitectural de sistem distribuit care implica un provider, un requester si un broker si foloseste tehnologii ca SOAP, WSDL si UDDI este specific pentru:

Arhitecturi bazate pe obiecte distribuite

Arhitecturi bazate pe servicii

Arhitecturi bazate pe evenimente

Arhitecturi bazate pe nivele (layered)

Care paradigma de calcul distribuit ofera resurse computationale clientilor “on demand” si “utility-based”?

Calculul orientat pe servicii

Cloud computing

Calculul privat

Calculul orientat pe obiecte distribuite

Care din urmatoarele sunt proprietati a doua evenimente concurente din punct de vedere logic:

Trebuie sa apara in acelasi moment de timp fizic

Nu trebuie sa se afecteze reciproc

Pot sa apara in momente de timp fizic diferite

Exista o cale (path) in diagrama spatiu – timp ce conecteaza cele doua evenimente

Nu exista o cale (path) in diagrama spatiu – timp ce conecteaza cele doua evenimente

Considerati tranzactia T1 ce se executa pe obiectul O1 si tranzactia T2 ce se executa pe alt obiect O2. Care din urmatoarele nu trebuie sa fie considerate ca operatii conflictuale in controlul concurentei

T1 citeste / T2 citeste

T1 scrie / T2 scrie

T1 scrie / T2 citeste

T1 citeste / T2 scrie

AJAX este o tehnica ce se utilizeaza pentru:

Gazduirea de stub-uri a procedurilor remote in RPC

Gazduirea de stub-uri ale procedurilor remote in RMI

Imbunatatirea interactiunii client-server pe Web

Imbunatatirea operatiunilor de cautare a serviciilor de catre broker

In algoritmul de snapshot Chandy-Lamport, Marker-ul este utilizat pentru a decide despre (bifati toate raspunsurile corecte):

Ce mesaje pot fi incluse in snapshot

Cand algoritmul isi termina executia

Politica utilizata pentru asamblarea starii globale

In care sistem de timp logic, doua sau mai multe evenimente din procese diferite pot sa aiba timestamp-uri identice?

Timp Scalar

Timp Vertical

Ambele

Situatia in care o tranzactie citeste o valoare ce a fost produsa de o alta tranzactie care in final se intrerupe (aborts) este numita:

Intrerupere in cascada

Update pierdut

Transactions deadlock

Returnari inconsistente (inconsistent retrieval)

Asigurarea proprietatilor de izolare si consistenta in executia tranzactiilor concurente este sarcina carei din urmatoarele componente:

Transaction manager

Scheduler

Data manager

Care din urmatoarele afirmatii e adevarata despre fragmentarea pe verticala a unui tabel in design-ul unei baze de date distribuita?

Toate fragmentele trebuie sa includa coloanele chei primare a tabelului original

Doar un singur fragment trebuie sa includa coloanele chei primare a tabelului original

Fragmentele pot sa nu includa coloanele chei primare a tabelului original

Vertical fragmentation must allow table re-construction from the fragments

Which of the following statements about the scalability of distributed systems are true?

A scalable system can handle an increasing number of users simultaneously

A load balancer can be used in horizontal scaling to distribute requests to loaded frontend servers

Vertical scaling can be implemented by adding more backend servers to a system

Vertical scaling can be achieved by adding more CPUs to a server

Which of the following statements about micro services design is true?

Microservices offer high cohesion and low coupling

Microservices offer low cohesion and high coupling

Which of the following statements about the NodeJs server are not true?

Uses a main thread for processing the blocking events -> (correct: non-blocking)

Uses more worker threads for non-blocking events

Uses a main thread for processing the non-blocking

Processes events asynchronously

Which of the following statements about the Network Time Protocol (NTP) are correct?

Uses hardware based time-stamping

Uses server organized in a stratum model on the internet

Uses atomic reference clocks

Uses scalar timestamping

Which of the following statements about global states in distributed systems are not true?

In a strongly consistent global state, all messages recorded as received are recorded as sent and the channels are not empty

For an inconsistent global state, the messages cross the cut in a time space diagram from the past to the future

For an inconsistent global state, the messages cross the cut in a time space diagram from the future to the past

A consistent global space can be captured using Singhal-Kshemkalyani Differential Technique

Which of the following statements is true about data distribution?

Increases availability

Increases consistency

An inter-process communication where the sender is blocked until an acknowledge message arrives, and the receiving application is not necessary to be active:

Persistent synchronous communication

Transient synchronous communication

Persistent asynchronous communication

Transient asynchronous communication

Which of the following statements represents false assumptions about a Distributed System?

The topology is fixed and does not change

The network is heterogeneous

Network is not reliable and messages may be lost

Communication bandwidth is infinite

Which type of virtualization uses the technique for the binary translation of instructions?

Para-virtualization

Full virtualization using a type 1 hypervisor

Hardware-assisted virtualization

Full virtualization using a type 2 hypervisor

Which of the following are true about vertical fragmentation of a table in distributed storage design? (mai sus da in ro)

Only one of the fragments must include the primary key column(s) of the original table

All fragments must include the primary key column(s) of the original table.

The fragments may not include the primary key column(s) of the original table.

Vertical fragmentation must allow table re-construction from the fragments

For which type of logical clocks the event timestamp must include the processor index for having totally ordered events in a distributed system? (o mai fost)

Scalar time

Both

Vector time

Which of the following are true about atomic commit protocols in distributed transactions?

Each Participant must call the join() operation defined by the Coordinator interface

The Coordinator must call the join() operation defined by Participants interface

In a distributed transaction, the Coordinator must have the references to all Participants

Check all true statements about Virtual Machines (VMs).

All VMs running on the same server must have the same guest Operating System

VMs can be replicated in multiple servers

VM instances may be created from a template VM

The virtualization layer provides services for data exchange between VMs;

Which of the following are specific only to distributed deadlocks?

Deadlock situations are detected by using wait-for graphs

Deadlock situations are detected by using timeout timers

A phantom deadlock situation may happen during a deadlock detection process

Concurrency control is the main task of (o mai fost)

Recovery Manager

Data Manager

Scheduler

Transaction Manager

Which of the following is the correct definition of the reliable one to one communication:

Any message in the outgoing message buffer is delivered to the incoming message buffer

The message received is identical to one sent, and no messages are delivered twice

Any message in the outgoing message buffer is delivered to the incoming message buffer and the message received is identical to one sent, and no messages are delivered twice

The transparency type that offers the same interface to a service, no matter if it is local or remote is called:

Access transparency

Location transparency

Migration transparency

Concurrency transparency

Check all true statements about a Global Conceptual Schema (GCS)

Includes how distributed tables are fragmented

Includes storage locations for distributed tables

Includes information about tables replication

Includes all primary keys of the tables

Includes all foreign keys of the tables

Google App Engine (GAE) is a type of:

Infrastructure as a Service

Software as a Service

Platform as a Service

None of the indicated answers

Failure is easier to be detected for synchronous distributed systems than for asynchronous distributed systems

False

True

In which type of virtualization technique the guest operating system (GOS) needs to be modified to be adapted to the hypervisor?

Full Virtualization of type 1

Full Virtualization of type 2

Para-Virtualization

Hardware-assisted Virtualization;

False) In a distributed system, strong consistency of replicated data has a positive impact over latency.

True

False

Check all true statements about a RPC client stub

It should be compiled and linked with the client code

It should be compiled and linked with the server code

It should implement the remote procedure

It is used to achieve the goal of semantic transparency

Check all true statements about a binding agent used in client - server binding process

Binding agent location should be known by both servers and clients

Binding agent location should be known only by the clients

Communication overhead is low when most of the clients are short lived

Replicated binding agents involves high overhead

Replicated binding agents involves low overhead

An interprocess communication where both send and receive operations are blocking operations is called:

Transient communication

Persistent communication

Synchronous communication

Asynchronous communication

Which of the following statements are true about the Chandy-Lamport snapshot algorithm?

It works for all types of communication networks

It works only for specific communication networks

The marker must be sent by a node before taking its snapshot

The marker must be sent by a node after taking its snapshot

A node may send the marker before or after taking its snapshot

Which of the followings are true about table fragmentation in distributed storage design:

Table fragments must not be replicated

Table fragments must be stored at different sites

Table fragmentation increases data processing parallelism

Table fragmentation helps in case of disaster recovery

Which of the following are true about nested distributed transactions?

Children sub-transactions may run on different machines;

Children sub-transactions must run on the same machine;

Children sub-transactions may run in parallel;

For consistency reasons, children sub-transactions must be executed one after the

Other;

Consider a transaction T operating on an object x and transaction U operating on an object y. Check all true statements below.

T read / U read is a conflicting situation;

T read / U write is a conflicting situation;

T write/ U read is a conflicting situation;

T write / U write is a conflicting situation;

The mentioned operations are not conflicting situations;

Check the true statement.

Asynchronous distributed system failures can be detected by using timeout timers;

Synchronous distributed system failures can be detected by using timeout timers;

Timeout timers can be used for failure detection in both synchronous and asynchronous distributed systems

Which of the following is true about a flat distributed transaction?

Each transaction operation must be completed before executing another operation;

Each transaction operation must be executed at a different site;

Each transaction operation must be completed before executing another operation;

Transaction operations may be executed in parallel;

The unwanted situation that may happen, when the acquired locks are progressively released before transaction execution completion is called:

Lost update situation

Inconsistent retrieval situation

Cascading abort situation

Deadlock situation

Which of the following is true about the events in a distributed system?

The timestamp of an event represents the physical time when the event happened;

Concurrent logical events must occur at the same physical time

Logically concurrent events are not causally related events;

Which ACID property describes the fact that all transaction effects should be saved into permanent storage after the successful transaction commit.

Consistency

Isolation

Durability

Atomicity

Which of the following are features of Optimistic Concurrency Control (OCC) technique?

It is recommended for private workspace transaction implementations;

It is recommended for write-ahead log transaction implementations;

There are no conflicts in OCC, hence all concurrent transactions may commit without any validation;

In OCC all transactions must be validated before they are allowed to commit.

Which of the following represent benefits of resource virtualization?

Server consolidation

Elastic usage of hardware resource

Avoids single points of failure

Limited infrastructure and resource demand

Data replication improves data consistency.

True

False

The ability of a distributed system to limit the latency due to system faults is called:

Reliability

Consistency

Availability

Efficiency

Which of the following statements are true about the atomic Two-Phase Commit Protocol (2PC) for Distributed Transactions?

The participants must save the involved transaction data into permanent storage before voting commit;

The participants must save the involved transaction data into permanent storage after voting commit;

Each protocol step defines a timeout timer for avoiding algorithm blocking;

The coordinator may decide to commit the transaction even if a participant voted abort

Which ACID property describes the fact that the execution of a transaction is transparent to other transactions until the successful completion of the commit action, even if the transactions run in parallel:

Atomicity

Consistency

Durability

Isolation

{"name":"Ds exam", "url":"https://www.quiz-maker.com/QPREVIEW","txt":"Which of the following statements represent false assumptions about a Distributed System?, Which of the following statements are true about time-space diagrams?, Which of the following objectives should be achieved by concurrency control Algorithms?","img":"https://www.quiz-maker.com/3012/images/ogquiz.png"}