Load Balancing and Auto Scaling Questions and Answers

11.

Assertion (A): Session affinity is best avoided where it can be.
Reason (R): It ties users to particular backends and so unbalances the pool and complicates scaling in.

A.A is true but R is false
B.A is false but R is true
C.Both A and R are true, and R is the correct explanation of A
D.Both A and R are true, but R is not the correct explanation of A

Answer: C

Both are true and the reason gives both objections, and it also means the loss of one backend loses the sessions it was holding, so a restart during a rolling deploy logs those users out. Every deployment then becomes visible to users who were midway through a task.

Discuss in Forum

12.

An auto scaling policy based on average processor utilisation will

A.add instances on a fixed schedule set in advance
B.add instances but never remove any of them again
C.scale on the time of day and on nothing else
D.add instances when the average rises above a threshold

Answer: D

Scheduled scaling is a separate mechanism for known patterns. Choosing the right metric matters, since processor use is a poor proxy for load in a memory or input output bound service. The metric should be whichever resource actually saturates first under load.

Discuss in Forum

13.

Consider the following statements about scaling metrics.
1. Processor utilisation suits compute bound workloads.
2. Queue length or request latency may suit others better.
3. Processor utilisation is the correct metric for every workload.

A.1 and 2 only
B.3 only
C.1 only
D.All of them

Answer: A

The first two are correct. The third is wrong, and scaling a memory bound service on processor use produces a system that never scales when it actually needs to. Queue depth is often the honest signal where processor use is a misleading one.

Discuss in Forum

14.

A cooldown period after a scaling action exists to

A.stop the metric being collected while capacity changes
B.let the new capacity take effect before deciding again
C.delay the scaling action itself by a fixed interval
D.reduce the hourly cost of the instances in the group

Answer: B

Without it the group reacts to a metric that has not yet responded and overshoots. Flapping between scaling out and in is the usual symptom of too short a cooldown. Each cycle costs launch time and money without ever settling at the right size.

Discuss in Forum

15.

Assertion (A): Without a cooldown an auto scaling group can oscillate between adding and removing capacity.
Reason (R): It reacts to a metric that has not yet reflected the previous action.

A.A is true but R is false
B.A is false but R is true
C.Both A and R are true, and R is the correct explanation of A
D.Both A and R are true, but R is not the correct explanation of A

Answer: C

Both are true and the reason gives the mechanism, and the oscillation costs money and destabilises the service at the same time, and it is entirely avoidable by widening the thresholds. A gap of twenty points between scale out and scale in is a common starting value.

Discuss in Forum

16.

State whether the following is true or false.
Flapping in an auto scaling group is usually a sign that the thresholds or cooldown need adjusting.

A.False
B.True only at large scales
C.False only at small scales
D.True

Answer: D

Scale out and scale in thresholds that are too close together cause it, as does too short a cooldown. Widening the gap between them is the usual first remedy. Lengthening the evaluation window before a decision has much the same effect.

Discuss in Forum

17.

Instance warm-up time matters to auto scaling because

A.a new instance takes time before it can serve traffic
B.an instance once launched never becomes ready at all
C.the load balancer ignores instances added after startup
D.it changes the hourly billing rate for the whole group

Answer: A

Scaling begun when load is already at the limit arrives too late if warm-up takes minutes. Scaling on a leading indicator rather than a lagging one is the usual answer. Otherwise every scaling decision arrives one warm-up period behind the traffic.

Discuss in Forum

18.

Assertion (A): Auto scaling should be triggered on a leading rather than a lagging indicator where possible.
Reason (R): New capacity takes time to become useful, so waiting for saturation is too late.

A.A is false but R is true
B.Both A and R are true, and R is the correct explanation of A
C.Both A and R are true, but R is not the correct explanation of A
D.A is true but R is false

Answer: B

Both are true and the reason gives the timing problem, and queue depth or request rate often lead processor saturation by exactly the interval needed. Choosing the metric well matters more than tuning the thresholds around it.

Discuss in Forum

19.

Instances take 3 minutes to become ready and traffic doubles in 2 minutes. What follows for a reactive scaling policy?

A.no scaling is needed for a rise of this speed
B.the cooldown period prevents any problem here
C.capacity arrives after the peak has passed
D.capacity arrives in good time to meet the peak

Answer: C

Reacting after the rise begins cannot beat a warm-up longer than the rise itself. Pre-warming, predictive scaling or a larger baseline are the available answers. None of them is free, so the choice depends on how costly a brief overload is.

Discuss in Forum

20.

All of the following help an auto scaling group respond to a sharp spike EXCEPT

A.predictive or scheduled scaling ahead of the event
B.a larger minimum capacity as a baseline
C.shortening the instance warm-up time
D.lengthening the cooldown period considerably

Answer: D

A longer cooldown makes the group slower to respond, which is the opposite of what a spike needs. The other three all bring capacity forward in time. A spike is won or lost on how early capacity starts, not on how fast it scales.

Discuss in Forum

Loading…

Topics

Other Subjects

Tags: Load Balancing and Auto Scaling Questions and Answers || Load Balancing and Auto Scaling MCQ Questions and Answers || Load Balancing and Auto Scaling GK Questions and Answers || Load Balancing and Auto Scaling GK MCQ Questions || Load Balancing and Auto Scaling Multiple Choice Questions and Answers || Load Balancing and Auto Scaling GK || GK on Load Balancing and Auto Scaling || Cloud Computing Questions and Answers || Cloud Computing MCQ Questions and Answers || Cloud Computing GK Questions and Answers || GK on Cloud Computing

Load Balancing and Auto Scaling Questions and Answers

Topics

Other Subjects

About Us

Stay informed about all important Dates & Notifications