QUESTION $1$

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect ones.$)$

To achieve predictable latency and horizontal scaling the following must be TRUE:

$1.$

You achieve scale by denormalization, minimizing the need to JOIN data across tables.

$2.$

By limiting the types of queries you can run, you can achieve faster and predictable performance.

$3.$

You do not need to rewrite your data to support different access patterns

$4.$

None of the above

$5$ points

QUESTION $2$

Which of the following statements are TRUE about EMR $($Amazon Elastic Map Reduce Service$)?$

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect ones.$)$

$1.$

EMR can be configured to provide Effiency and Infinite Scale

$2.$

EMR supports ALT $($Aggregator Leaf Trailer Architecture$)$

$3.$

EMR can be configured ot provide Pattern Flexibility and Efficiency

$4.$

EMR can be configured to provide Pattern Flexibility and Infinite Scale

$5.$

EMR in conjunction with S$3$ can offer polygot persistence in certain use cases

$5$ points

QUESTION $3$

Which one of the following statements is TRUE about data consistency?

$1.$

Eventual consistency is a form of weak consistency

$2.$

You can achieve weak consistency through a quorum

$3.$

For certain type of data consistency strategies updates are distributed to replicas in a lazy fashion

$4.$

All of the above

$5$ points

QUESTION $4$

Which of the following is true about a narrow Spark transformation?

$1.$

The data required to compute the results of the trasformation resides on multiple partitions.

$2.$

The data required to compute the results of the transformation resides on a single partition.

$3.$

Both of the above are correct statements.

$4.$

None of the above are correct statements.

$5$ points

QUESTION $5$

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect answer.$)$

Given what we have learned in the course, select the statements that are TRUE?

$1.$

Kafka can be used for Queue, Workers and Databases

$2.$

Workers and Databases can be represented by AWS Glue, Apache Spark and S$3$

$3.$

Workers and Databases can be represented by a combination of Spark, Spark Streaminng, Hive, Hive Catalog

$4.$

Diagram above can be implemented using either Lambda or Kappa architecture approaches

$5.$

Diagram represents a logical construct for Online Analytical Processing $($OLAP$)$ solution

$5$ points

QUESTION $6$

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect ones$)$

Given a peer$-$to$-$peer replication system with replication factor of $5,$ Which of the following statements are TRUE?

$1.$

At least $3$ replicas have to agree on writes to achieve write consistency.

$2.$

At least $3$ nodes contacted for a single read and $3$ nodes agree on writes to achieve strong read consistency

$3.$

Both $1)$ and $2)$ needs to be true to achieve Tunable AP system with a replication factor of $3.$

$4.$

None of the above

$5$ points

QUESTION $7$

Is Data Lake an example of polygot persistence?

True

False

$5$ points

QUESTION $8$

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect answer.$)$

A web application emits multiple types of events for operational reporting. Critical events must be captured immediately before processing can continue, but informational events do not need to deplay processing.

$1.$

You can use Kappa Architecture but not Lambda Architecture to solve for the above scenario

$2.$

Capture critical events in Kafka

$3.$

Process informational events using Spark Streaming

$4.$

You can store both critical and informational events into Kafka and use Spark streaming across two separate consumer groups to solve for the above scenario.

$5$ points

QUESTION $9$

Given a distributed system that is spread across machines on two racks. Which of the following statements are TRUE?

$($Choose the options carefully. Please note you will get partial credit for correct responses but also will be penalized for incorrect ones.$)$

$1.$

Client can connect to machines in either racks but for some reason machines on either racks cannot talk to each other. In such a situation CP system will lock out requests until connection between racks is restored and all the replicas can be updated

$2.$

Client can connect to machines in either racks but for some reason machines on either racks cannot talk to each other. A system that enforces AP will honor all requests but cannot guarantee that all the replicas will be updated

$3.$

Client can connect to machines in either rack