[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"cheat-sheet---en":3,"domain-info---en":3,"topic-info----en":3,"lesson-aws-certified-cloudops-engineer-associate-monitoring-logging-performance-event-driven-remediation-automated-remediation-patterns-en":4,"prev-aws-certified-cloudops-engineer-associate-monitoring-logging-performance-event-driven-remediation-automated-remediation-patterns-en":19,"next-aws-certified-cloudops-engineer-associate-monitoring-logging-performance-event-driven-remediation-automated-remediation-patterns-en":32},null,{"locked":5,"reason":6,"meta":7,"item":3},true,"paywall",{"title":8,"description":9,"isFree":10,"estimatedMinutes":11,"difficulty":12,"learningObjectives":13},"Automated Remediation Patterns","Assemble detection and repair into working self-healing designs: alarm actions, event-driven runbooks, AWS Config remediation, scheduled sweeps, the guardrails each one needs, and where AWS DevOps Agent and Kiro fit.",false,24,"advanced",[14,15,16,17,18],"Identify which of the four detection sources a remediation scenario is describing","Choose the lowest-capability remediation pattern that can actually perform the required repair","Configure an AWS Config rule with an automatic remediation action and explain its retry semantics","Apply the guardrails that keep an automated fix from causing a larger incident","Explain what AWS DevOps Agent and Kiro contribute to an incident response loop and where the human gate stays",{"locked":5,"reason":6,"meta":20,"item":3},{"title":21,"description":22,"isFree":10,"estimatedMinutes":23,"difficulty":12,"learningObjectives":24},"Systems Manager Automation Runbooks","Write and run Systems Manager Automation runbooks: the action vocabulary, step properties that make a repair safe to run unattended, the two-role permission chain, approvals, and rate control across a fleet.",25,[25,26,27,28,29,30,31],"Explain what makes a runbook different from a script and identify its document type and schema version","Select the right Automation action for a step from the available action types","Use maxAttempts, timeoutSeconds, onFailure, onCancel, and isCritical to make a runbook safe to run unattended","Trace the permission chain from the caller through AutomationAssumeRole to the target service","Diagnose the three classic Automation permission failures from their error messages","Run a runbook across a fleet with targets, concurrency, and an error threshold","Start a runbook from an EventBridge rule with the correct target configuration",{"locked":10,"reason":3,"meta":33,"item":44},{"title":34,"description":35,"isFree":5,"estimatedMinutes":11,"difficulty":36,"learningObjectives":37},"EC2 Right-Sizing and Compute Optimizer","Find the instances that are the wrong size using Compute Optimizer findings, the metrics CloudWatch cannot see on its own, and the credit model that makes T instances behave unlike every other family.","intermediate",[38,39,40,41,42,43],"Explain why CPUUtilization alone cannot tell you whether an instance is the right size","Interpret Compute Optimizer finding classifications, finding reasons, and performance risk","Calculate how long a burstable instance can sustain a given CPU level before it exhausts its credits","Distinguish standard mode from unlimited mode and predict the behavior of each at zero credits","Choose between Compute Optimizer, Cost Explorer right sizing, and Trusted Advisor for a given question","Run a right-sizing change on a live instance without losing data or configuration",{"id":45,"title":34,"body":46,"description":35,"difficulty":36,"estimatedMinutes":11,"extension":705,"infographics":706,"isFree":5,"learningObjectives":724,"meta":725,"navigation":5,"path":726,"quiz":727,"seo":795,"stem":796,"__hash__":797},"courses/courses/aws-certified-cloudops-engineer-associate/en/domains/01-monitoring-logging-performance/04-compute-and-storage-performance/01-ec2-right-sizing-and-compute-optimizer.md",{"type":47,"value":48,"toc":696},"minimark",[49,62,70,75,80,83,89,96,100,103,106,113,123,129,142,146,149,199,262,280,287,291,294,305,312,328,333,336,354,368,371,399,402,406,409,412,548,551,573,587,590,594,597,653,656,660,689,692],[50,51,52,53,57,58,61],"p",{},"You inherit an account with 40 web servers on ",[54,55,56],"code",{},"m5.2xlarge",". Every instance is healthy. No alarm has fired in eight months. CloudWatch shows ",[54,59,60],{},"CPUUtilization"," averaging 8% with peaks around 22%. The monthly compute bill is roughly $11,000, and about $8,000 of that is buying capacity nobody uses.",[50,63,64,65,69],{},"Nothing here is broken, which is exactly why nobody looks at it. Right-sizing is the part of the CloudOps job with no incident attached to it, and the exam tests it because the skills involved are the same ones you use when something ",[66,67,68],"em",{},"is"," broken: reading utilization metrics correctly, knowing which metric is missing, and knowing which tool answers which question.",[71,72,74],"h2",{"id":73},"cpu-utilization-is-one-dimension-of-four","CPU utilization is one dimension of four",[50,76,77,79],{},[54,78,60],{}," is the metric everyone reaches for, and on its own it is close to useless for sizing decisions.",[50,81,82],{},"An instance has at least four capacity dimensions that can independently run out: CPU, memory, network, and storage I/O. CloudWatch publishes instance metrics for three of them out of the box. It does not publish memory, because the hypervisor cannot see inside the guest. To AWS, your operating system's memory is opaque; only an agent running inside the instance can report how much of it is in use.",[50,84,85,86,88],{},"This produces a specific and very common failure shape: an application that is slow while ",[54,87,60],{}," sits at 12%. The instance is starved for memory, the operating system is swapping, and every graph in the console looks calm. If a scenario describes low CPU and poor performance, memory is the first thing to suspect, and the fix that unblocks the diagnosis is installing the unified CloudWatch agent so that memory becomes a metric you can see.",[50,90,91,92,95],{},"The same blind spot affects the tooling built on top of these metrics. Compute Optimizer analyzes memory utilization ",[66,93,94],{},"only"," for resources with the CloudWatch agent installed. Without it, you get findings based on CPU, network, and I/O, and the service will happily recommend a smaller instance for a workload that is actually memory-bound.",[71,97,99],{"id":98},"what-compute-optimizer-does","What Compute Optimizer does",[50,101,102],{},"AWS Compute Optimizer reads the configuration and CloudWatch utilization metrics of your resources and returns a recommended configuration, with the projected utilization if you adopted it.",[50,104,105],{},"Four facts about how it works shape most exam questions:",[50,107,108,112],{},[109,110,111],"strong",{},"You must opt in."," It does nothing until you activate it for a standalone account, a member account, or the management account of an organization. A scenario where \"Compute Optimizer shows no recommendations\" often ends at this.",[50,114,115,118,119,122],{},[109,116,117],{},"The default lookback is 14 days."," It analyzes the last 14 days of metrics and refreshes recommendations daily. Activating ",[109,120,121],{},"enhanced infrastructure metrics",", a paid recommendation preference, extends the window to 93 days. That is the answer whenever a workload has a monthly close, a quarterly batch, or any peak that a 2-week window would miss entirely.",[50,124,125,128],{},[109,126,127],{},"It needs enough data."," Recommendations can take up to 24 hours to appear after opt-in, and a resource with too little metric history gets no finding at all.",[50,130,131,134,135,138,139,141],{},[109,132,133],{},"It covers far more than EC2."," EC2 instances and Auto Scaling groups, EBS volumes, Lambda functions, ECS services on Fargate, RDS and Aurora databases, DynamoDB, ElastiCache, MemoryDB, DocumentDB, NAT gateways, WorkSpaces, SageMaker, and commercial software licenses. If a question asks which single service gives right-sizing recommendations across compute ",[66,136,137],{},"and"," storage ",[66,140,137],{}," databases, this is it.",[71,143,145],{"id":144},"reading-a-finding","Reading a finding",[50,147,148],{},"Every analyzed instance lands in one of three classifications:",[150,151,152,165],"table",{},[153,154,155],"thead",{},[156,157,158,162],"tr",{},[159,160,161],"th",{},"Finding",[159,163,164],{},"Meaning",[166,167,168,179,189],"tbody",{},[156,169,170,176],{},[171,172,173],"td",{},[109,174,175],{},"Under-provisioned",[171,177,178],{},"At least one specification does not meet the workload's requirements. Risk to performance.",[156,180,181,186],{},[171,182,183],{},[109,184,185],{},"Over-provisioned",[171,187,188],{},"At least one specification can be sized down, and nothing is under-provisioned. Risk to the bill.",[156,190,191,196],{},[171,192,193],{},[109,194,195],{},"Optimized",[171,197,198],{},"Every specification meets requirements and nothing is oversized. Compute Optimizer may still suggest a newer generation.",[50,200,201,202,205,206,209,210,213,214,217,218,217,221,217,224,217,227,217,230,217,233,217,236,239,240,243,244,247,248,243,251,254,255,243,258,261],{},"The classification tells you ",[66,203,204],{},"that"," something is wrong. The ",[109,207,208],{},"finding reason"," tells you ",[66,211,212],{},"what",", and it is specific: ",[54,215,216],{},"CPU over-provisioned",", ",[54,219,220],{},"Memory under-provisioned",[54,222,223],{},"EBS throughput under-provisioned",[54,225,226],{},"EBS IOPS over-provisioned",[54,228,229],{},"Network bandwidth under-provisioned",[54,231,232],{},"Network PPS under-provisioned",[54,234,235],{},"Disk IOPS",[54,237,238],{},"Disk throughput",", and their GPU equivalents. Each names the metric it was derived from, so you can go verify it yourself: EBS IOPS reasons come from ",[54,241,242],{},"VolumeReadOps"," and ",[54,245,246],{},"VolumeWriteOps",", network bandwidth from ",[54,249,250],{},"NetworkIn",[54,252,253],{},"NetworkOut",", network PPS from ",[54,256,257],{},"NetworkPacketsIn",[54,259,260],{},"NetworkPacketsOut",".",[50,263,264,265,268,269,217,272,275,276,279],{},"Note the split that catches people: ",[109,266,267],{},"Disk"," reasons refer to instance store volumes (",[54,270,271],{},"DiskReadOps",[54,273,274],{},"DiskWriteBytes","), while ",[109,277,278],{},"EBS"," reasons refer to attached EBS volumes. Different hardware, different fix. A disk finding is solved by changing the instance type; an EBS finding is often solved by modifying the volume, which the next lesson but one covers in detail.",[50,281,282,283,286],{},"Each recommendation also carries a ",[109,284,285],{},"performance risk"," from very low to very high (0 to 4 in the API). It is the maximum risk score across all the analyzed specifications, and it answers \"how likely is this smaller type to disappoint me?\" Very low means the type is predicted to always have enough capability. Anything higher is a prompt to test under real load first.",[71,288,290],{"id":289},"burstable-instances-follow-different-rules","Burstable instances follow different rules",[50,292,293],{},"Everything above assumes an instance can use 100% of its CPU whenever it wants. T instances cannot, and this is the single most-tested piece of EC2 sizing knowledge.",[50,295,296,297,300,301,304],{},"A burstable instance is sold with a ",[109,298,299],{},"baseline"," level of CPU it can sustain forever, plus a bucket of ",[109,302,303],{},"CPU credits"," that let it run above the baseline for a while. One CPU credit equals one vCPU running at 100% for one minute. Below baseline, the instance earns more than it spends and the bucket fills. Above baseline, it spends more than it earns and the bucket drains.",[50,306,307,308,311],{},"Here is a ",[54,309,310],{},"t3.large",", which has 2 vCPUs, earns 36 credits per hour, has a 30% baseline, and can accrue up to 864 credits (one full day of earning):",[313,314,315,319,322,325],"ul",{},[316,317,318],"li",{},"Demand rises to a steady 55% CPU.",[316,320,321],{},"Credits spent per minute = vCPUs x utilization = 2 x 0.55 = 1.1, so 66 per hour.",[316,323,324],{},"Credits earned per hour = 36.",[316,326,327],{},"Net drain = 30 credits per hour. Starting from a full 864, the balance hits zero in about 29 hours.",[329,330],"infographic",{"alt":331,"slug":332},"A two-panel chart showing a t3.large holding 55 percent CPU while its credit balance drains from 864 to zero over 29 hours, then diverging into a throttled standard-mode path and a billed unlimited-mode path.","t-instance-credit-burn",[50,334,335],{},"What happens at zero depends on the credit configuration mode:",[313,337,338,344],{},[316,339,340,343],{},[109,341,342],{},"Standard mode",": the instance is brought down to its baseline. CPU is clamped at 30% and the application slows, with no error, no alarm, and no obvious cause. This is the default for T2 and for T3 on a Dedicated Host.",[316,345,346,349,350,353],{},[109,347,348],{},"Unlimited mode",": the instance keeps running above baseline by spending ",[109,351,352],{},"surplus credits",". If its average CPU over a rolling 24 hours stays at or below baseline, the surplus is covered by the normal hourly price. If not, you pay a flat additional rate per vCPU-hour. This is the default for T3, T3a, and T4g.",[50,355,356,357,360,361,363,364,367],{},"The misconception to name out loud: people read \"burstable\" as a bonus and assume the instance can hold high CPU indefinitely. It cannot, in either mode, without a consequence. In standard mode the consequence is a performance cliff roughly a day after the load increased. In unlimited mode the consequence is a bill line that nobody was expecting. The metric that predicts both is ",[54,358,359],{},"CPUCreditBalance",", not ",[54,362,60],{},", so a T-instance fleet needs an alarm on the credit balance trending toward zero. ",[54,365,366],{},"CPUSurplusCreditsCharged"," is the one to watch for the billing side.",[50,369,370],{},"A few more credit facts worth holding:",[313,372,373,384,390,393],{},[316,374,375,376,379,380,383],{},"The accrual limit is always 24 hours' worth of earning. A ",[54,377,378],{},"t3.micro"," earns 12 per hour and accrues 288; a ",[54,381,382],{},"t3.2xlarge"," earns 192 per hour and accrues 4,608.",[316,385,386,387,389],{},"Baseline percentages are per vCPU and match what CloudWatch shows. A ",[54,388,310],{}," at baseline reads as 30% in the console.",[316,391,392],{},"For T3, T3a, and T4g, accrued credits survive a stop for 7 days. For T2 they are lost the moment you stop the instance.",[316,394,395,398],{},[109,396,397],{},"Launch credits"," exist only for T2 in standard mode. T3 and later launch in unlimited mode, so they can burst immediately without needing them.",[50,400,401],{},"If a T instance is consistently living above its baseline, it is the wrong instance for the workload. That is a right-sizing finding, not a tuning problem, and Compute Optimizer draws a burstable baseline line on its CPU graph so you can see it.",[71,403,405],{"id":404},"right-sizing-a-live-instance","Right-sizing a live instance",[50,407,408],{},"Once you have the finding, the change itself is mechanical, and the exam cares about the ordering and the side effects.",[50,410,411],{},"For an EBS-backed instance, changing the type requires the instance to be stopped:",[413,414,419],"pre",{"className":415,"code":416,"language":417,"meta":418,"style":418},"language-bash shiki shiki-themes material-theme-lighter github-light github-dark","aws ec2 stop-instances --instance-ids i-0abc123def4567890\naws ec2 wait instance-stopped --instance-ids i-0abc123def4567890\n\naws ec2 modify-instance-attribute \\\n  --instance-id i-0abc123def4567890 \\\n  --instance-type \"{\\\"Value\\\": \\\"m5.xlarge\\\"}\"\n\naws ec2 start-instances --instance-ids i-0abc123def4567890\n","bash","",[54,420,421,444,461,467,481,492,529,534],{"__ignoreMap":418},[422,423,426,430,434,437,441],"span",{"class":424,"line":425},"line",1,[422,427,429],{"class":428},"sbgvK","aws",[422,431,433],{"class":432},"s_sjI"," ec2",[422,435,436],{"class":432}," stop-instances",[422,438,440],{"class":439},"stzsN"," --instance-ids",[422,442,443],{"class":432}," i-0abc123def4567890\n",[422,445,447,449,451,454,457,459],{"class":424,"line":446},2,[422,448,429],{"class":428},[422,450,433],{"class":432},[422,452,453],{"class":432}," wait",[422,455,456],{"class":432}," instance-stopped",[422,458,440],{"class":439},[422,460,443],{"class":432},[422,462,464],{"class":424,"line":463},3,[422,465,466],{"emptyLinePlaceholder":5},"\n",[422,468,470,472,474,477],{"class":424,"line":469},4,[422,471,429],{"class":428},[422,473,433],{"class":432},[422,475,476],{"class":432}," modify-instance-attribute",[422,478,480],{"class":479},"s_hVV"," \\\n",[422,482,484,487,490],{"class":424,"line":483},5,[422,485,486],{"class":439},"  --instance-id",[422,488,489],{"class":432}," i-0abc123def4567890",[422,491,480],{"class":479},[422,493,495,498,502,505,508,511,513,516,518,521,523,526],{"class":424,"line":494},6,[422,496,497],{"class":439},"  --instance-type",[422,499,501],{"class":500},"sjJ54"," \"",[422,503,504],{"class":432},"{",[422,506,507],{"class":479},"\\\"",[422,509,510],{"class":432},"Value",[422,512,507],{"class":479},[422,514,515],{"class":432},": ",[422,517,507],{"class":479},[422,519,520],{"class":432},"m5.xlarge",[422,522,507],{"class":479},[422,524,525],{"class":432},"}",[422,527,528],{"class":500},"\"\n",[422,530,532],{"class":424,"line":531},7,[422,533,466],{"emptyLinePlaceholder":5},[422,535,537,539,541,544,546],{"class":424,"line":536},8,[422,538,429],{"class":428},[422,540,433],{"class":432},[422,542,543],{"class":432}," start-instances",[422,545,440],{"class":439},[422,547,443],{"class":432},[50,549,550],{},"The instance keeps its ID, its EBS volumes, its security groups, and its IAM role. Three things do not survive the stop and start:",[313,552,553,560,567],{},[316,554,555,556,559],{},"Data on ",[109,557,558],{},"instance store"," volumes, because the instance moves to different host hardware.",[316,561,562,563,566],{},"The ",[109,564,565],{},"public IPv4 address",", unless an Elastic IP is associated.",[316,568,562,569,572],{},[109,570,571],{},"CPU credit balance"," on a T2 instance.",[50,574,575,576,217,579,582,583,586],{},"Tags do the organizational half of this work. Right-sizing at fleet scale means being able to answer \"who owns this instance and what environment is it in\" before you resize it, which is why Skill 1.3.1 of the exam guide names resource tags alongside performance metrics. A consistent ",[54,577,578],{},"Environment",[54,580,581],{},"Owner",", and ",[54,584,585],{},"Application"," tag set is what turns a list of 400 findings into 12 conversations.",[50,588,589],{},"For instances in an Auto Scaling group, do not resize instances individually. Update the launch template with the new instance type and start an instance refresh, so the change is durable and the group replaces instances in a controlled way.",[71,591,593],{"id":592},"which-tool-answers-which-question","Which tool answers which question",[50,595,596],{},"Three services produce overlapping-sounding advice, and questions are written to separate them.",[150,598,599,612],{},[153,600,601],{},[156,602,603,606,609],{},[159,604,605],{},"Tool",[159,607,608],{},"What it is driven by",[159,610,611],{},"What it gives you",[166,613,614,627,640],{},[156,615,616,621,624],{},[171,617,618],{},[109,619,620],{},"Compute Optimizer",[171,622,623],{},"CloudWatch utilization metrics over 14 (or 93) days",[171,625,626],{},"A specific recommended type per resource, with projected utilization and performance risk",[156,628,629,634,637],{},[171,630,631],{},[109,632,633],{},"Cost Explorer rightsizing recommendations",[171,635,636],{},"The same Compute Optimizer engine, presented against your spend",[171,638,639],{},"Downsizing and idle-instance candidates with estimated savings, inside a cost view",[156,641,642,647,650],{},[171,643,644],{},[109,645,646],{},"Trusted Advisor",[171,648,649],{},"Fixed thresholds against a checklist",[171,651,652],{},"Flags for idle and underutilized instances, plus security, limits, and fault tolerance checks",[50,654,655],{},"The cue: \"recommend an instance type based on measured utilization\" is Compute Optimizer. \"Show me savings opportunities alongside my bill\" is Cost Explorer. \"Check my account against a list of best practices\" is Trusted Advisor.",[71,657,659],{"id":658},"exam-tips","Exam tips",[313,661,662,665,668,674,677,686],{},[316,663,664],{},"Low CPU plus a slow application means memory. The unblocking action is installing the CloudWatch agent, because memory is not a default CloudWatch metric and Compute Optimizer cannot analyze it without the agent.",[316,666,667],{},"\"Monthly batch job\", \"quarterly peak\", or \"seasonal load\" plus Compute Optimizer points at enhanced infrastructure metrics and its 93-day lookback. The default 14 days would never see the peak.",[316,669,670,671,673],{},"A T instance that gets slow after about a day of elevated load is a credit exhaustion question. The metric is ",[54,672,359],{},"; the fix is either unlimited mode (accepting the surplus charge) or a non-burstable instance family.",[316,675,676],{},"Watch which default the question implies. T2 defaults to standard mode and throttles; T3, T3a, and T4g default to unlimited and bill instead.",[316,678,679,680,682,683,685],{},"Do not confuse ",[109,681,267],{}," finding reasons (instance store) with ",[109,684,278],{}," finding reasons (attached volumes). They lead to different remediations.",[316,687,688],{},"Changing an instance type needs the instance stopped, and stopping it destroys instance store data and releases a non-Elastic public IP.",[50,690,691],{},"The habit to carry forward: before you change any instance's size, know which of its four capacity dimensions is actually constrained. CPU is the one you can see by default, memory is the one you have to go get, and for T instances the constraint is not a dimension at all but a credit balance. The next lesson takes the third dimension, network, and shows why an instance can be nowhere near its stated bandwidth and still be throttled.",[693,694,695],"style",{},"html pre.shiki code .sbgvK, html code.shiki .sbgvK{--shiki-light:#E2931D;--shiki-default:#6F42C1;--shiki-dark:#B392F0}html pre.shiki code .s_sjI, html code.shiki .s_sjI{--shiki-light:#91B859;--shiki-default:#032F62;--shiki-dark:#9ECBFF}html pre.shiki code .stzsN, html code.shiki .stzsN{--shiki-light:#91B859;--shiki-default:#005CC5;--shiki-dark:#79B8FF}html pre.shiki code .s_hVV, html code.shiki .s_hVV{--shiki-light:#90A4AE;--shiki-default:#005CC5;--shiki-dark:#79B8FF}html pre.shiki code .sjJ54, html code.shiki .sjJ54{--shiki-light:#39ADB5;--shiki-default:#032F62;--shiki-dark:#9ECBFF}html .light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html.light .shiki span {color: var(--shiki-light);background: var(--shiki-light-bg);font-style: var(--shiki-light-font-style);font-weight: var(--shiki-light-font-weight);text-decoration: var(--shiki-light-text-decoration);}html .default .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .shiki span {color: var(--shiki-default);background: var(--shiki-default-bg);font-style: var(--shiki-default-font-style);font-weight: var(--shiki-default-font-weight);text-decoration: var(--shiki-default-text-decoration);}html .dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}html.dark .shiki span {color: var(--shiki-dark);background: var(--shiki-dark-bg);font-style: var(--shiki-dark-font-style);font-weight: var(--shiki-dark-font-weight);text-decoration: var(--shiki-dark-text-decoration);}",{"title":418,"searchDepth":463,"depth":463,"links":697},[698,699,700,701,702,703,704],{"id":73,"depth":446,"text":74},{"id":98,"depth":446,"text":99},{"id":144,"depth":446,"text":145},{"id":289,"depth":446,"text":290},{"id":404,"depth":446,"text":405},{"id":592,"depth":446,"text":593},{"id":658,"depth":446,"text":659},"md",[707],{"slug":332,"concept":708,"style":709,"aspectRatio":710,"labels":711},"A two-panel time-series chart sharing one x-axis of hours, showing what happens to a t3.large held above its baseline. The top panel plots CPU utilization as a flat line at 55 percent with a dashed horizontal line at the 30 percent baseline and the gap between them shaded as the burn zone. The bottom panel plots the CPU credit balance falling in a straight line from the 864 accrual limit to zero at hour 29, with two divergent continuations after that point: standard mode drops CPU to the baseline, unlimited mode keeps CPU at 55 percent and starts a shaded surplus area that is billed. The emphasis is that the CPU graph looks healthy the entire time while the credit graph is the one that predicts the incident.","diagram","16:9",[712,713,714,715,716,717,718,719,720,721,722,723],"t3.large: 2 vCPUs, earns 36 credits per hour, accrual limit 864","Top panel: CPU utilization (%)","Sustained demand: 55%","Baseline: 30%","Burn zone: every minute above baseline spends credits","Bottom panel: CPUCreditBalance","Start: 864 credits (full bucket)","Net drain: 30 credits per hour (66 spent, 36 earned)","Hour 29: balance reaches zero","Standard mode: CPU is clamped to the 30% baseline, the application slows","Unlimited mode: CPU stays at 55%, surplus credits accrue and are billed per vCPU-hour","CPUUtilization looks healthy right up to the cliff. CPUCreditBalance is the metric that warns you.",[38,39,40,41,42,43],{},"/courses/aws-certified-cloudops-engineer-associate/en/domains/01-monitoring-logging-performance/04-compute-and-storage-performance/01-ec2-right-sizing-and-compute-optimizer",{"passingScore":728,"questions":729},70,[730,739,747,755,761,771,779,787],{"question":731,"type":732,"options":733,"correctAnswer":737,"explanation":738},"An EC2 instance shows a sustained CPUUtilization of 12% and no CloudWatch alarms have ever fired, yet the application on it is slow. Compute Optimizer reports the instance as under-provisioned. What is the most likely finding reason?","single",[734,735,736,737],"CPU under-provisioned, because CPUUtilization is averaged across vCPUs","The instance has no finding reason, because under-provisioned findings require CPU over 80%","Network PPS under-provisioned, which always accompanies low CPU","Memory under-provisioned, identified from the memory metric published by the CloudWatch agent","The hypervisor can see CPU, network, and disk activity, but memory pressure lives inside the guest operating system, so it reaches CloudWatch only through the agent. Low CPU with a slow application is the classic memory-bound shape, and Compute Optimizer names it explicitly once the agent is publishing memory data.",{"question":740,"type":732,"options":741,"correctAnswer":744,"explanation":746},"How far back does Compute Optimizer look when it analyzes utilization metrics, by default?",[742,743,744,745],"24 hours","93 days","14 days","30 days","The default analysis window is the last 14 days of CloudWatch metrics. Extending it to 93 days requires activating enhanced infrastructure metrics, which is a paid recommendation preference, so a scenario about a workload with a monthly or quarterly peak points at that feature rather than the default.",{"question":748,"type":732,"options":749,"correctAnswer":750,"explanation":754},"A t3.large runs at a steady 55% CPU utilization starting from a full credit balance of 864. It earns 36 credits per hour and its baseline is 30%. Roughly how long can it hold 55% before the balance is empty?",[750,751,752,753],"About 29 hours","About 12 hours","About 6 hours","It never empties, because 55% is below 100%","Credits spent per minute equal vCPUs times utilization, so 2 vCPUs at 55% spend 1.1 credits per minute, or 66 per hour, against 36 earned. The net drain of 30 credits per hour against an 864 balance gives roughly 29 hours, which is why this failure mode surfaces a full day after the traffic change that caused it.",{"question":756,"type":732,"options":757,"correctAnswer":759,"explanation":760},"True or False: a T3 instance launched with default settings will slow to its baseline CPU once its credit balance reaches zero.",[758,759],"True","False","T3, T3a, and T4g launch in unlimited mode by default, so at zero credits they spend surplus credits and keep running above baseline, billed at a flat rate per vCPU-hour. The clamp to baseline is standard mode behavior, which is the default only for T2 and for T3 on a Dedicated Host.",{"question":762,"type":763,"options":764,"correctAnswers":769,"explanation":770},"Which statements about Compute Optimizer are correct? (Choose 2.)","multiple",[765,766,767,768],"It generates recommendations automatically for every account with no configuration","You must opt in before it analyzes your resources","Memory utilization is analyzed only for resources with the unified CloudWatch agent installed","It only covers EC2 instances and EC2 Auto Scaling groups",[766,767],"Compute Optimizer requires an explicit opt-in at the account or organization management account level, and it cannot see memory without the agent. It also covers far more than EC2, including EBS volumes, Lambda functions, ECS services on Fargate, RDS and Aurora databases, DynamoDB, ElastiCache, and NAT gateways.",{"question":772,"type":732,"options":773,"correctAnswer":776,"explanation":778},"A finance team wants a report of which specific instances can be downsized and by how much, based on measured utilization rather than spend patterns. Which tool answers this most directly?",[774,775,776,777],"AWS Cost Explorer rightsizing recommendations","AWS Trusted Advisor cost optimization checks","AWS Compute Optimizer","AWS Budgets with a usage threshold","Compute Optimizer is the utilization-driven engine: it analyzes CloudWatch metrics and returns a specific recommended instance type with projected utilization and a performance risk score. Cost Explorer rightsizing surfaces the same underlying recommendations inside a cost view, and Trusted Advisor flags idle and underutilized resources against fixed thresholds without proposing a target type.",{"question":780,"type":732,"options":781,"correctAnswer":784,"explanation":786},"You need to change a running m5.2xlarge to an m5.xlarge. Which sequence is correct for an EBS-backed instance?",[782,783,784,785],"Modify the instance type in place while it is running, then reboot","Create an AMI, launch a new instance from it, then terminate the original","Stop the instance, change the instance type, then start it","Detach the root volume, resize it, then reattach it to a new instance","Changing the instance type of an EBS-backed instance requires the stopped state, and the instance keeps its EBS volumes, instance ID, and configuration through the change. Note that a stop and start moves the instance to new host hardware, so any instance store volumes are lost and the public IPv4 address changes unless an Elastic IP is attached.",{"question":788,"type":732,"options":789,"correctAnswer":791,"explanation":794},"Compute Optimizer reports a recommended instance type with a performance risk of very high. What does that mean?",[790,791,792,793],"The recommendation will save the most money of the three options offered","The recommended type is predicted to be the most likely of the options to fall short of the workload's requirements","The recommendation cannot be applied without a support case","The current instance type is failing and must be replaced immediately","Performance risk scores the likelihood that an instance type does not meet the workload's requirements, on a scale from very low to very high (0 to 4 in the API), taking the maximum score across CPU, memory, EBS, disk, and network specifications. A very high score is a signal to validate under real load before migrating, not a statement about the current instance's health.",{"title":34,"description":35},"courses/aws-certified-cloudops-engineer-associate/en/domains/01-monitoring-logging-performance/04-compute-and-storage-performance/01-ec2-right-sizing-and-compute-optimizer","99UphY-7bWUo3kTKTRo3DaitXn1ZswebERl5BWavS38"]