Which statement correctly distinguishes Amazon EC2 Dedicated Hosts from Dedicated Instances?
Intermediate · Choose one
Explanation
Answer: D. Both options give single-tenant hardware, but only Dedicated Hosts expose the physical server's sockets, cores, and host ID. That visibility is what allows customers to bring licenses that are counted per socket, per core, or per VM. Dedicated Hosts are billed per host and add host affinity and targeted instance placement. Dedicated Instances are billed per instance, support a narrower set of BYOL scenarios such as SQL Server with License Mobility, and offer no placement control.
A. This reverses the billing models. Dedicated Hosts carry a per-host charge; Dedicated Instances carry a per-instance charge plus a per-Region fee for dedicated tenancy.
B. This reverses the two. Dedicated Instances are the option that provides isolation without hardware visibility.
C. Host affinity and targeted instance placement are Dedicated Hosts features only. Dedicated Instances do not support them, so price is not the only difference.
62
Which TWO statements accurately describe Amazon EC2 Spot Instances? (Choose TWO.)
Intermediate · Select 2 answers
Explanation
Answer: B and C. Spot Instances draw on unused Amazon EC2 capacity, which is why AWS can price them at up to 90% below On-Demand. The price of that discount is reclaimability: when AWS needs the capacity for On-Demand or committed-use customers, it interrupts the Spot Instance after a two-minute warning. Understanding both halves together, deep discount and interruptibility, is what lets a candidate decide whether a workload belongs on Spot.
A. Spot prices adjust gradually based on supply and demand for spare capacity, so the rate is not fixed for the life of the instance.
D. Spot requires no term commitment at all. Commitments of 1 or 3 years belong to Reserved Instances and Savings Plans.
E. Reserving capacity in a specific Availability Zone is the function of a zonal Reserved Instance or an On-Demand Capacity Reservation. Spot provides the opposite: capacity that AWS may take back.
63
A company moves log files to S3 Standard-Infrequent Access as soon as they are written, then deletes most of them after 10 days. Its S3 bill rises rather than falls. What is the most likely cause?
Intermediate · Choose one
Explanation
Answer: A. S3 Standard-IA carries a 30-day minimum storage duration charge. An object deleted or transitioned before 30 days is still billed as though it had been stored for the full 30 days, so short-lived data placed in an infrequent-access class costs more than leaving it in S3 Standard. Minimum duration charges also apply to S3 One Zone-IA at 30 days, S3 Glacier Instant Retrieval and Glacier Flexible Retrieval at 90 days, and S3 Glacier Deep Archive at 180 days, and they are a leading cause of lifecycle policies that backfire.
B. Request charges do differ by storage class, but for a workload of this shape the dominant and documented effect is the minimum storage duration charge, not the per-request cost of writing.
C. Lifecycle rules work with S3 Standard-IA. Transitioning an object changes its storage class rather than creating a second copy.
D. S3 does not replicate to another Region unless the customer explicitly configures replication. Standard-IA stores objects across at least three Availability Zones within one Region.
64
A company has a large dataset whose access pattern is unpredictable and changes over time. It wants to reduce storage cost automatically without risking retrieval charges when access spikes. Which Amazon S3 storage class should it use?
Intermediate · Choose one
Explanation
Answer: A. S3 Intelligent-Tiering exists precisely for unknown or changing access patterns. It monitors object access and moves objects between a Frequent Access tier, an Infrequent Access tier after 30 days without access, and an Archive Instant Access tier after 90 days, then moves them straight back on access. Critically, AWS charges no retrieval fees for Intelligent-Tiering, so an unexpected access spike does not produce a surprise bill. The only additional charge is a small monthly monitoring and automation fee per object.
B. S3 Standard-IA charges a per-GB retrieval fee, so an access spike on data the company assumed was cold produces exactly the surprise charge it wants to avoid. It also carries a 30-day minimum storage duration.
C. Glacier Flexible Retrieval requires an explicit restore before objects can be read, imposes retrieval charges, and has a 90-day minimum storage duration, none of which suits unpredictable access.
D. One Zone-IA lowers cost by reducing durability protection to a single Availability Zone, and it still charges retrieval fees. It addresses a different trade-off entirely.
65
An application server in one Availability Zone communicates constantly with a database in a different Availability Zone in the same AWS Region. The team wants to reduce the resulting data transfer charges. Which statement should guide the redesign?
Intermediate · Choose one
Explanation
Answer: B. AWS charges for data transfer between Availability Zones in the same Region, and the charge applies to both the inbound and outbound side of the transfer. Communication between resources in the same Availability Zone over private IPv4 addresses is not charged. That gives the architect a real cost lever, though it has to be weighed against resilience, since collapsing tiers into one Availability Zone removes the fault isolation that multi-AZ deployment provides.
A. Reversing the direction changes nothing, because the charge applies to inbound and outbound traffic for the resources involved.
C. Cross-Availability Zone traffic within a Region is charged, so it is a legitimate and often significant line item for chatty application tiers.
D. Using public IPv4 or Elastic IP addresses makes matters worse, not better. Traffic routed through those addresses is charged even between resources in the same Region, because it leaves through the internet gateway.