Page

$2$

of $2$

Consider an international airline which operates both in passenger segment and cargo segment.

For every flight that is flying, the airline capture a lot of data in real time which can be used for

live tracking of flight status, modelling the flight schedules as well as for preventive maintenance

schedule etc. Also at the same time, the same data was used for various analytical purposes which

are oriented towards improving the airline operations and also for the predicting the passenger

loads, cargo loads in near future and devising the marketing strategies around it$.$ Identify the

appropriate data processing architecture that can help in achieving these use cases. With a help of

architectural diagram, represent the proposed system architecture.

Answer:

Its case of Lambda architecture as data needs to be processed in batch mode as well as in real time

mode.

Identification of the architecture type

Drawing each of the following system component explaining scenarios for the processing

across batch and streaming layers

$$ Source data representation

$$ Batch layer

$$ Speed layer

$$ Serving layer

$$ Query layer

A gas filling system where LPG cylinders get filled. Streaming application sends the filling data

such as pressure, timestamp, status $($OK$/$NOT OK$)$ to the downstream for subsequent analysis.

The system$$s working and functionality is monitored towards potential leakage of gas.

Discuss data delivery model message delivery semantics

Data delivery model to be looked into $4$ dimensions:

$1.$ What would be the impact on filling station if the communication between the collection tier and analysis

tier interrupted for extended period of time? There is element of life risk involved in case the system failure.

I can afford to lose couple of messages

$2.$ How much duration worth of data $($in terms of cylinders and time stamps$)$ that filling system can tolerate

losing. Data of few minutes can be managed if it is lost. Because the monitoring systems have robust

clearance systems in place.

$3.$ Would you anticipate that filling system need to store historical data? It requires some data to be stored in

order to infer on new events. Down sampling is a technique one can use for accurate results.

Explain the $$Split brain problem$$ and its handling mechanism in distributed systems?

Loss of connectivity between two groups of systems is split brain problem.

$$ To enter in degraded state $$ disallow changes to distributed state until connectivity is restored

$$ Allow one of partitions to remain fully functional while degrading the capabilities of other

partition.