5 exercises on infrastructure key phrases. Choose the most natural and professional option.
0 / 5 completed
1 / 5
Your application is running out of RAM on the current cloud instance and performance is degrading. Which phrase best describes the situation in a technical discussion?
"We're hitting the memory ceiling on this instance" is the precise technical phrase for RAM exhaustion on a specific compute instance. "Memory ceiling" is widely understood in cloud and ops discussions — it pinpoints RAM as the bottleneck and "this instance" scopes it to the current machine, implying an upgrade or horizontal scaling is needed. Option A ("server is broken") is vague and alarmist. Option B ("slow today") is imprecise and doesn't identify the root cause. Option D ("database is down") is a completely different problem. Precise problem naming is critical in infrastructure discussions to steer the team towards the right solution.
2 / 5
Your team is over-provisioned — you're paying for large instances but utilisation is consistently under 20%. What recommendation do you make?
"I'd recommend right-sizing the cluster" is the standard cloud cost optimisation phrase. "Right-sizing" means matching instance type and count to actual resource consumption — moving to smaller or fewer instances when utilisation is low. It is a formal recommendation category used by AWS, GCP, and Azure cost management tools. Option B ("add more servers") is the opposite of what's needed and increases waste. Option C ("scale up") means moving to larger instances, which again increases cost. Option D ("load balancer") solves a different problem — distributing traffic, not reducing over-provisioned capacity.
3 / 5
A stakeholder asks why last month's cloud bill was higher than expected. You need to explain the spend breakdown. Which phrase opens that discussion correctly?
"The cost breakdown is..." is the professional phrase for presenting a structured, line-item analysis of cloud expenditure. It signals that you have data organised by service, resource type, or environment (e.g., compute, storage, data transfer). This phrase sets up a clear and analytical conversation. Option A ("it's complicated") is evasive and loses stakeholder confidence. Option B ("overspent on everything") is imprecise and alarming without actionable detail. Option C ("billing is automatic") is factually true but completely dodges the question. Cloud cost governance requires precise, data-driven language — "breakdown" is the right term.
4 / 5
You are presenting a cost optimisation proposal that involves switching from on-demand cloud instances to a committed-use pricing model. Which phrase best summarises the saving?
"We could save X% by moving to reserved instances" is the standard phrase for presenting Reserved Instance (RI) or Committed Use Discount (CUD) proposals. It is quantified, action-oriented, and uses the correct cloud-native terminology. Reserved instances (AWS), committed use discounts (GCP), and reserved capacity (Azure) all refer to this pricing model — paying upfront or committing for 1-3 years in exchange for significant discounts (often 30-70%). Options A and C are problem statements, not solutions. Option D (vendor negotiation) is a separate strategy and does not describe the RI model. Always quantify savings proposals with a percentage or dollar figure.
5 / 5
During a multi-region deployment discussion, you need to justify choosing a specific AWS/GCP region for your primary workload. Which statement is most professionally precise?
"This region has better latency for our users" is the correct technical justification for region selection from a user-experience perspective. Latency — the round-trip time between a user's device and the server — is the primary performance metric for region placement decisions. It is measurable, user-centric, and directly tied to business impact. Option A ("newer") is irrelevant to performance or cost decisions. Option B ("popular") is not a technical criterion. Option D ("more services") is sometimes valid but does not speak to user-facing performance. In infrastructure discussions, latency data (e.g., from cloudping.info or traceroute) is the gold standard for region justification.