[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"cheat-sheet---en":3,"domain-info---en":3,"topic-info----en":3,"lesson-cloud-computing-fundamentals-cloud-services-and-architecture-cloud-service-models-infrastructure-as-a-service-en":4,"prev-cloud-computing-fundamentals-cloud-services-and-architecture-cloud-service-models-infrastructure-as-a-service-en":253,"next-cloud-computing-fundamentals-cloud-services-and-architecture-cloud-service-models-infrastructure-as-a-service-en":506},null,{"locked":5,"reason":3,"meta":6,"item":17},false,{"title":7,"description":8,"isFree":9,"estimatedMinutes":10,"difficulty":11,"learningObjectives":12},"Infrastructure as a Service (IaaS)","The service model that hands you raw compute, storage, and networking and leaves the operating system, runtime, and application on your side of the line, with AWS EC2, Azure Virtual Machines, and Google Compute Engine as the products that deliver it.",true,15,"beginner",[13,14,15,16],"Define Infrastructure as a Service (IaaS) using the NIST framework","Identify which layers of the stack IaaS leaves for the customer versus the provider","Name IaaS products from major providers and what each one delivers","Explain why virtualized infrastructure still leaves real operational work on the customer",{"id":18,"title":7,"body":19,"description":8,"difficulty":11,"estimatedMinutes":10,"extension":193,"infographics":194,"isFree":9,"learningObjectives":195,"meta":196,"navigation":9,"path":197,"quiz":198,"seo":250,"stem":251,"__hash__":252},"courses/courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/01-infrastructure-as-a-service.md",{"type":20,"value":21,"toc":181},"minimark",[22,27,31,35,38,41,45,124,127,131,154,157,161,164,167,171,174,178],[23,24,26],"h2",{"id":25},"the-layer-a-hypervisor-hands-you","The layer a hypervisor hands you",[28,29,30],"p",{},"You already know a hypervisor can slice one physical server into dozens of isolated virtual machines, and that a real cloud service lets you request one of those machines yourself, on demand, with no ticket and no wait. What you have not seen yet is what a provider actually sells you once that machine exists, and how much of the stack above it lands in your lap instead of theirs. Infrastructure as a Service is the model that hands you the most: raw compute, storage, and networking, with the operating system and everything above it left for you to install, configure, and patch.",[23,32,34],{"id":33},"the-nist-definition-in-practice","The NIST definition, in practice",[28,36,37],{},"NIST's definition of IaaS is precise about the split: the consumer can \"provision processing, storage, networks, and other fundamental computing resources\" and \"deploy and run arbitrary software, which can include operating systems and applications.\" The consumer does not manage the underlying cloud infrastructure, but has control over operating systems, storage, and deployed applications, and possibly limited control of select networking components, such as host firewalls.",[28,39,40],{},"That last clause matters. You do not touch the physical switch or the hypervisor, but you do configure the firewall rules that decide which traffic reaches your own virtual machine. IaaS gives you a box, full control of everything inside it, and a narrow window into the network path that reaches it.",[23,42,44],{"id":43},"who-manages-what","Who manages what",[46,47,48,61],"table",{},[49,50,51],"thead",{},[52,53,54,58],"tr",{},[55,56,57],"th",{},"Layer",[55,59,60],{},"Who manages it in IaaS",[62,63,64,73,80,87,95,102,109,116],"tbody",{},[52,65,66,70],{},[67,68,69],"td",{},"Networking (physical)",[67,71,72],{},"Provider",[52,74,75,78],{},[67,76,77],{},"Servers and storage hardware",[67,79,72],{},[52,81,82,85],{},[67,83,84],{},"Virtualization (the hypervisor)",[67,86,72],{},[52,88,89,92],{},[67,90,91],{},"Operating system",[67,93,94],{},"You",[52,96,97,100],{},[67,98,99],{},"Runtime and middleware",[67,101,94],{},[52,103,104,107],{},[67,105,106],{},"Application code",[67,108,94],{},[52,110,111,114],{},[67,112,113],{},"Data",[67,115,94],{},[52,117,118,121],{},[67,119,120],{},"Host firewall rules",[67,122,123],{},"You (limited networking control)",[28,125,126],{},"Compare that to the 5 characteristics from the previous topic: resource pooling and rapid elasticity are what let the provider hand you that virtual machine in seconds. IaaS is where those characteristics get packaged into an actual product you can buy.",[23,128,130],{"id":129},"real-products-that-deliver-iaas","Real products that deliver IaaS",[132,133,134,142,148],"ul",{},[135,136,137,141],"li",{},[138,139,140],"strong",{},"Virtual machines:"," AWS EC2 (Elastic Compute Cloud), Azure Virtual Machines, Google Compute Engine",[135,143,144,147],{},[138,145,146],{},"Block and object storage:"," AWS EBS and S3, Azure Managed Disks and Blob Storage, Google Persistent Disk and Cloud Storage",[135,149,150,153],{},[138,151,152],{},"Virtual networking:"," AWS VPC, Azure Virtual Network, Google VPC",[28,155,156],{},"Every one of these gives you a resource you provision, configure, and are responsible for keeping patched and running. None of them ships an application. They ship the pieces you build one from.",[23,158,160],{"id":159},"worked-example-what-launching-a-vm-actually-commits-you-to","Worked example: what launching a VM actually commits you to",[28,162,163],{},"Launch an EC2 instance and AWS asks 2 questions first: which OS image (an AMI, for example Amazon Linux 2023) and which instance type. Pick a t3.micro, 2 vCPUs and 1 GiB of memory, and AWS's on-demand rate in US East (N. Virginia) is $0.0104 an hour, about $7.59 for a 730-hour month, with storage and data transfer billed separately.",[28,165,166],{},"That price buys you a running Linux box and nothing else. Nobody at AWS installs a web server on it, applies its security patches, or configures its firewall. You do all 3, the same way you would on a physical machine, except you never touched a data center, ordered hardware, or waited for a shipment.",[23,168,170],{"id":169},"the-misconception-its-in-the-cloud-does-not-mean-its-someone-elses-problem","The misconception: \"it's in the cloud\" does not mean \"it's someone else's problem\"",[28,172,173],{},"It is tempting to hear \"cloud\" and assume the provider now handles security and maintenance for you. IaaS does not work that way. The provider took the procurement, power, cooling, and physical maintenance off your plate, the exact burden the first lesson in this course described. It did not take the operating system off your plate. An unpatched IaaS virtual machine is just as vulnerable as an unpatched physical server sitting in your own building; the vulnerability is still yours to close.",[23,175,177],{"id":176},"where-this-leaves-you","Where this leaves you",[28,179,180],{},"IaaS gives you the most control of the 3 service models, and, in exchange, the most responsibility: everything from the operating system up is yours to run. The next lesson covers what happens when a provider takes the operating system and runtime off your plate too, and what you get, and give up, in return.",{"title":182,"searchDepth":183,"depth":183,"links":184},"",3,[185,187,188,189,190,191,192],{"id":25,"depth":186,"text":26},2,{"id":33,"depth":186,"text":34},{"id":43,"depth":186,"text":44},{"id":129,"depth":186,"text":130},{"id":159,"depth":186,"text":160},{"id":169,"depth":186,"text":170},{"id":176,"depth":186,"text":177},"md",[],[13,14,15,16],{},"/courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/01-infrastructure-as-a-service",{"passingScore":199,"questions":200},70,[201,210,216,224,234,242],{"question":202,"type":203,"options":204,"correctAnswer":207,"explanation":209},"A team needs to install a specific Linux kernel version required by a legacy driver, with full root access to configure it. Which service model gives them that?","single",[205,206,207,208],"SaaS","Traditional on-premises hosting only","IaaS","PaaS","IaaS is the only cloud service model that hands the customer the operating system layer, including root access to install a specific kernel version. PaaS and SaaS both take the OS off the customer's hands, and on-premises hosting is not a cloud service model at all, just the alternative the cloud replaces.",{"question":211,"type":203,"options":212,"correctAnswer":214,"explanation":215},"In Infrastructure as a Service, the provider manages the guest operating system running inside your virtual machine.",[213,214],"True","False","In IaaS, the customer manages the guest operating system, along with the runtime, applications, and data. The provider's responsibility stops at the hypervisor: it manages the physical servers, storage, networking, and virtualization layer that the virtual machine runs on.",{"question":217,"type":203,"options":218,"correctAnswer":221,"explanation":223},"Under the NIST definition of IaaS, which of these might the consumer have limited control over, in addition to the operating system, storage, and deployed applications?",[219,220,221,222],"Nothing, IaaS and PaaS manage identical layers","The physical network switches","Select networking components, such as host firewalls","The hypervisor itself","NIST's IaaS definition specifically calls out possible limited control of select networking components, like host firewalls, alongside full control of the OS, storage, and applications. The physical switches and the hypervisor stay firmly on the provider's side in every case.",{"question":225,"type":226,"options":227,"correctAnswers":232,"explanation":233},"Which of the following are the customer's responsibility under IaaS? (Select all that apply.)","multiple",[228,229,230,231],"Patching the operating system","Maintaining the physical server hardware","Installing and configuring the application runtime","Managing the hypervisor",[228,230],"Patching the OS and installing the runtime both sit above the hypervisor, which is exactly where the customer's responsibility begins in IaaS. Physical hardware maintenance and hypervisor management stay with the provider in every one of the 3 service models, not just IaaS.",{"question":235,"type":203,"options":236,"correctAnswer":239,"explanation":241},"Which of these is an IaaS product?",[237,238,239,240],"Salesforce","Google Workspace","Google Compute Engine","AWS Elastic Beanstalk","Google Compute Engine provisions virtual machines and leaves the OS, runtime, and application to the customer, which is the IaaS pattern. Salesforce and Google Workspace are complete applications you just use, SaaS, and Elastic Beanstalk manages the OS and runtime for you, which makes it PaaS.",{"question":243,"type":203,"options":244,"correctAnswer":246,"explanation":249},"A team runs a t3.micro EC2 instance priced at $0.0104 per hour continuously for a 730-hour month, before counting storage or data transfer. Approximately what does the compute cost that month?",[245,246,247,248],"$0.76","$7.59","$75.90","$759.00","$0.0104 multiplied by 730 hours comes to about $7.59. That figure only covers the compute hours themselves; a real bill would add storage and any data transfer on top of it.",{"title":7,"description":8},"courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/01-infrastructure-as-a-service","E2TtoH5uaTA_XFqzAN-GojxuORCUsRcIDLg93LcegCI",{"locked":5,"reason":3,"meta":254,"item":264},{"title":255,"description":256,"isFree":5,"estimatedMinutes":257,"difficulty":258,"learningObjectives":259},"Total Cost of Ownership","Why the number on a cloud invoice, or a hardware quote, is never the full cost of running a workload, and how to compare cloud and on-premises costs honestly over several years.",18,"intermediate",[260,261,262,263],"Define total cost of ownership (TCO) and distinguish it from the price on a single invoice or hardware quote","Identify the direct and indirect cost categories a full TCO comparison has to include on both the cloud and on-premises sides","Explain what FinOps is and why it exists as an ongoing practice rather than a one-time calculation","Evaluate a scenario to judge whether a workload's TCO favors cloud or on-premises infrastructure",{"id":265,"title":255,"body":266,"description":256,"difficulty":258,"estimatedMinutes":257,"extension":193,"infographics":437,"isFree":5,"learningObjectives":446,"meta":447,"navigation":9,"path":448,"quiz":449,"seo":503,"stem":504,"__hash__":505},"courses/courses/cloud-computing-fundamentals/en/domains/01-cloud-concepts-and-economics/03-cloud-economics/03-total-cost-of-ownership.md",{"type":20,"value":267,"toc":428},[268,272,275,278,282,285,289,292,295,299,302,305,309,312,399,402,405,410,414,417,420,422,425],[23,269,271],{"id":270},"the-invoice-that-only-tells-you-the-price-of-the-meter","The invoice that only tells you the price of the meter",[28,273,274],{},"In 2022, 37signals, the company behind Basecamp and the email service Hey, decided its AWS and Google Cloud bills had become one of the most expensive parts of running the business, and began moving that infrastructure back into its own data centers. By 2024, its annual cloud bill had dropped from $3.2 million to $1.3 million, a reported $2 million a year in savings, after an upfront hardware purchase of around $700,000, with the company projecting more than $10 million saved over 5 years. That is a real, well-documented case of owning hardware costing less than renting the equivalent capacity, for that company's specific, steady, predictable workload.",[28,276,277],{},"It is also, on its own, an incomplete comparison, and understanding why is exactly what total cost of ownership is for.",[23,279,281],{"id":280},"what-tco-actually-means","What TCO actually means",[28,283,284],{},"Total cost of ownership is the complete cost of acquiring, running, and maintaining a system over a defined period, usually 3 to 5 years, not just the price of any single line item. A cloud invoice shows the meter: compute, storage, and data transfer, billed for the month just used. A hardware quote shows the sticker price of the box. Neither one is TCO. TCO is what you get once you add everything that surrounds that visible number on both sides of the comparison.",[23,286,288],{"id":287},"direct-costs-versus-the-costs-nobody-puts-on-the-invoice","Direct costs versus the costs nobody puts on the invoice",[28,290,291],{},"The direct, metered costs are the ones a cloud bill already shows: compute, storage, and data transfer. Data transfer deserves a specific note, because it is not symmetric. AWS, like most providers, does not charge to move data into its cloud, but does charge, on a tiered basis, to move data out, called egress. A workload that reads a small amount of data but exports large amounts of it regularly can carry a much bigger data-transfer bill than its compute costs alone would suggest.",[28,293,294],{},"Everything else is an indirect cost, and it sits on both sides of the comparison, not just one. On the cloud side: the one-time work of migrating or re-architecting a workload to run there, staff time spent configuring, monitoring, and optimizing cloud resources, the cost-management and monitoring tooling bought specifically to track that spend, and security or compliance tooling layered on top. On the on-premises side: power and cooling for the room the hardware sits in, the real estate itself, a hardware refresh every few years as equipment ages out, and staff time spent racking, patching, and eventually disposing of retired equipment, the same maintenance burden the first lesson in this domain described as one of the original reasons the cloud exists.",[23,296,298],{"id":297},"the-trap-comparing-only-whats-easy-to-compare","The trap: comparing only what's easy to compare",[28,300,301],{},"It is tempting to run this comparison by looking up 2 numbers, this quarter's cloud bill and the sticker price of equivalent server hardware, and declaring whichever one is smaller the winner. Neither number is TCO. The cloud bill leaves out the staff and tooling built up around it. The hardware sticker price leaves out power, cooling, and the refresh cycle that arrives again in 3 to 5 years.",[28,303,304],{},"The 37signals case is a useful, real example of how easy this trap is to fall into even when a company is genuinely trying to be rigorous. Independent reporting on the move specifically noted that its published savings figures did not yet account for future hardware refresh costs, the additional operations staff the move required, or the ongoing cost of running data center space, power, and cooling. That is not a criticism of 37signals's decision, which may well still hold up once those categories are counted. It is a demonstration that even a company publicly making the case for owning hardware had not, at the point those figures were reported, finished counting every category a full TCO comparison requires.",[23,306,308],{"id":307},"worked-example-a-3-year-comparison-done-honestly","Worked example: a 3-year comparison, done honestly",[28,310,311],{},"Take a workload with steady, predictable traffic and compare 3 years of on-premises ownership against 3 years in the cloud, counting the indirect costs on both sides.",[46,313,314,327],{},[49,315,316],{},[52,317,318,321,324],{},[55,319,320],{},"Cost category",[55,322,323],{},"On-premises (3 years)",[55,325,326],{},"Cloud (3 years)",[62,328,329,340,351,361,371,382],{},[52,330,331,334,337],{},[67,332,333],{},"Compute and storage / hardware",[67,335,336],{},"$120,000 (upfront purchase)",[67,338,339],{},"$270,000 ($7,500/month)",[52,341,342,345,348],{},[67,343,344],{},"Power and cooling",[67,346,347],{},"$45,000 ($15,000/year)",[67,349,350],{},"--",[52,352,353,356,358],{},[67,354,355],{},"Cost-management and monitoring tooling",[67,357,350],{},[67,359,360],{},"$6,000 ($2,000/year)",[52,362,363,366,368],{},[67,364,365],{},"One-time migration",[67,367,350],{},[67,369,370],{},"$20,000",[52,372,373,376,379],{},[67,374,375],{},"Staff time",[67,377,378],{},"$150,000 ($50,000/year)",[67,380,381],{},"$75,000 ($25,000/year)",[52,383,384,389,394],{},[67,385,386],{},[138,387,388],{},"Total",[67,390,391],{},[138,392,393],{},"$315,000",[67,395,396],{},[138,397,398],{},"$371,000",[28,400,401],{},"For this specific, steady workload, the honest 3-year comparison favors on-premises hardware, by about $56,000, and not simply because the cloud's metered bill is bigger than the hardware's sticker price. It only becomes clear once staffing and tooling are counted on both sides. This is the same shape of decision 37signals made, at a much larger scale.",[28,403,404],{},"Change one assumption and the answer can flip. Give this workload a traffic pattern that spikes to 3 times its baseline for 2 weeks a year, and the on-premises side now needs hardware sized for a peak it uses for 2 weeks and leaves idle the rest of the year, the exact overprovisioning problem the cloud was built to remove. The cloud side, by contrast, simply scales up for those 2 weeks and back down afterward, paying only for the extra capacity it actually used. That single change in workload shape is often enough to tip the comparison the other way. Neither \"the cloud is always cheaper\" nor \"the cloud is always a ripoff\" is a rule you can memorize; TCO is a calculation you run for a specific workload, not a verdict that transfers from one company's case to another's.",[406,407],"infographic",{"alt":408,"slug":409},"An iceberg diagram showing the visible cloud invoice above the waterline and the larger, hidden costs of staff, tooling, migration, power, and cooling below it.","total-cost-of-ownership-iceberg",[23,411,413],{"id":412},"finops-the-discipline-that-keeps-tco-from-going-stale","FinOps: the discipline that keeps TCO from going stale",[28,415,416],{},"A TCO calculation is a snapshot, priced once for a decision made at one point in time. Cloud spending does not stay still after that decision: it changes every time an engineer spins up a new resource, which is exactly what makes it different from a fixed annual data center budget set once and left alone. FinOps is the discipline built to manage that difference. The FinOps Foundation defines it as an operational framework and cultural practice that maximizes the business value of technology, enables timely, data-driven decisions, and creates financial accountability through collaboration between engineering, finance, and business teams, rather than treating cost as something only a finance department reviews after the fact.",[28,418,419],{},"This is the same idea AWS's own Well-Architected Framework points at with its cost optimization principles: implementing cloud financial management as a real organizational capability, and continuously analyzing and attributing expenditure so a workload's owner can see, and act on, what it actually costs. FinOps is what it looks like to run that analysis continuously instead of once a year.",[23,421,177],{"id":176},[28,423,424],{},"A cloud invoice and a hardware quote are both partial answers. Before comparing 2 ways of running a workload, list every direct and indirect cost on both sides, price them over the same multi-year window, and only then compare totals. Get that habit right once, in a spreadsheet, and FinOps is simply what it looks like to keep doing it continuously instead of once.",[28,426,427],{},"That closes out the economics of the cloud: what problem it solves, how you pay for it, and what it actually costs to run. The next domain leaves cost behind and turns to the services themselves, starting with the 3 ways cloud providers package what they sell: IaaS, PaaS, and SaaS.",{"title":182,"searchDepth":183,"depth":183,"links":429},[430,431,432,433,434,435,436],{"id":270,"depth":186,"text":271},{"id":280,"depth":186,"text":281},{"id":287,"depth":186,"text":288},{"id":297,"depth":186,"text":298},{"id":307,"depth":186,"text":308},{"id":412,"depth":186,"text":413},{"id":176,"depth":186,"text":177},[438],{"slug":409,"concept":439,"style":440,"aspectRatio":441,"labels":442},"An iceberg diagram split by a waterline. Above the waterline, small and clearly visible, sits an invoice icon labeled with the direct, metered costs: compute, storage, data transfer. Below the waterline, a much larger submerged mass carries the indirect costs: staff time, monitoring and cost-management tooling, security and compliance tooling, migration effort, and, on a mirrored on-premises side, power, cooling, and hardware refresh cycles. A footer strip carries the takeaway that TCO counts everything below the waterline too.","diagram","4:3",[443,444,445],"Above the waterline: the invoice. Compute, storage, and data transfer, the price everyone sees.","Below the waterline: migration, staff time, monitoring and security tooling, power, cooling, and hardware refresh cycles.","Total cost of ownership counts everything below the waterline too.",[260,261,262,263],{},"/courses/cloud-computing-fundamentals/en/domains/01-cloud-concepts-and-economics/03-cloud-economics/03-total-cost-of-ownership",{"passingScore":199,"questions":450},[451,459,463,471,479,487,495],{"question":452,"type":203,"options":453,"correctAnswer":455,"explanation":458},"What does total cost of ownership (TCO) measure that a monthly cloud invoice does not?",[454,455,456,457],"TCO measures only the compute portion of a bill","TCO measures the complete cost of running a system, including the indirect costs an invoice or hardware quote leaves out","TCO measures only the upfront hardware price","TCO is another name for the on-demand pricing rate","A monthly invoice or a hardware quote each shows one visible price. TCO adds everything around that price, staff time, tooling, migration effort, power and cooling, over a multi-year period, so the comparison reflects the full cost rather than just the part that happens to appear on a single bill.",{"question":460,"type":203,"options":461,"correctAnswer":214,"explanation":462},"37signals' publicly reported cloud repatriation savings, according to independent reporting, fully accounted for hardware refresh costs, new operations staff, and data center power and cooling.",[213,214],"Independent coverage of 37signals' move noted that its published savings figures did not yet include future hardware refresh costs, the new operations roles the move required, or the ongoing cost of running its own data center space. That gap is exactly the kind of incomplete comparison a rigorous TCO analysis is built to catch, even when the company making the comparison is genuinely trying to be transparent.",{"question":464,"type":203,"options":465,"correctAnswer":468,"explanation":470},"A finance team compares 2 options for a workload by looking up this quarter's cloud bill and the sticker price of equivalent server hardware, then picks whichever number is lower. What is the main problem with this comparison?",[466,467,468,469],"It ignores the on-demand discount percentage","It correctly identifies the cheaper option in every case","It leaves out the indirect costs on both sides, staff time, tooling, power and cooling, and migration, that a single invoice or sticker price does not show","Cloud bills and hardware prices cannot be compared to each other at all","Both numbers in this comparison are partial. The cloud bill leaves out the staff and tooling built up around it; the hardware sticker price leaves out power, cooling, and the refresh cycle that arrives again in 3 to 5 years. A real TCO comparison prices every category on both sides over the same window before declaring a winner.",{"question":472,"type":226,"options":473,"correctAnswers":477,"explanation":478},"Which of the following are indirect costs that belong in a cloud TCO calculation, based on this lesson? (Select all that apply.)",[474,475,476,355],"Staff time spent configuring and monitoring cloud resources","The on-demand hourly compute rate itself","Migration and re-architecting work needed to move a workload to the cloud",[474,476,355],"Staff time, migration effort, and the tooling used to watch cloud spend are all indirect costs that sit around the metered bill, not on it. The on-demand hourly rate is the one direct, metered cost in this list, the part of TCO that already shows up on the invoice.",{"question":480,"type":203,"options":481,"correctAnswer":484,"explanation":486},"Using the 3-year worked example in this lesson, what is the main reason the on-premises total ends up lower than the cloud total for that specific workload?",[482,483,484,485],"The cloud path has a larger one-time migration cost than the entire on-premises hardware purchase","On-premises power and cooling costs are cheaper than any cloud data-transfer fee","The workload is steady and predictable, so it gets no benefit from the cloud's ability to scale usage up and down","Cloud providers charge more for storage than for compute in every case","The cloud's cost advantage usually comes from matching spend to a workload that varies. A workload that runs at the same steady size around the clock never uses that flexibility, so it pays the cloud's metered rate every hour without ever getting the payoff that variability would provide, which is what lets the on-premises total come out ahead here.",{"question":488,"type":203,"options":489,"correctAnswer":490,"explanation":494},"A workload that normally runs at a steady baseline spikes to 3 times its usual size for 2 weeks a year. How does this change a TCO comparison between on-premises and cloud?",[490,491,492,493],"It can favor the cloud, since on-premises would need hardware sized for the rare spike, while cloud capacity can scale up temporarily and back down","It does not change anything; TCO comparisons never depend on the shape of a workload","It favors on-premises further, since owned hardware handles spikes better than the cloud does","It eliminates the need to calculate TCO at all","On-premises hardware has to be sized for the worst 2 weeks of the year and then sits underused the rest of the time, exactly the overprovisioning problem from earlier in this domain. Cloud capacity can scale up for those 2 weeks and back down afterward, so a spiky workload often tips a TCO comparison in the cloud's favor even when a steady one does not.",{"question":496,"type":203,"options":497,"correctAnswer":499,"explanation":502},"According to the FinOps Foundation, what is FinOps primarily about?",[498,499,500,501],"Cutting cloud costs to the lowest possible number, regardless of business impact","A cultural practice and operational framework that creates financial accountability and data-driven decisions across engineering, finance, and business teams","A pricing model offered directly by AWS, Azure, and Google Cloud","A once-a-year budgeting exercise kept separate from engineering","FinOps is defined by its own foundation as a cross-team cultural practice, not simply a cost-cutting mandate or a once-a-year budget review. It exists because cloud spending changes every time an engineer provisions a resource, so managing it well requires the same continuous attention as security or reliability, with finance and engineering making decisions together.",{"title":255,"description":256},"courses/cloud-computing-fundamentals/en/domains/01-cloud-concepts-and-economics/03-cloud-economics/03-total-cost-of-ownership","wol5owsWNcpygHYTaqOtd77uM9jYLnRLVOafyyyEKPU",{"locked":5,"reason":3,"meta":507,"item":515},{"title":508,"description":509,"isFree":5,"estimatedMinutes":10,"difficulty":11,"learningObjectives":510},"Platform as a Service (PaaS)","The service model that takes the operating system and runtime off your plate, so you push code and the platform runs it, with AWS Elastic Beanstalk, Heroku, and Google App Engine as the products that deliver it.",[511,512,513,514],"Define Platform as a Service (PaaS) using the NIST framework","Identify which layers move from customer to provider between IaaS and PaaS","Contrast deploying the same application through IaaS and through PaaS","Explain the boundary between PaaS and the serverless model covered later in this course",{"id":516,"title":508,"body":517,"description":509,"difficulty":11,"estimatedMinutes":10,"extension":193,"infographics":667,"isFree":5,"learningObjectives":668,"meta":669,"navigation":9,"path":670,"quiz":671,"seo":714,"stem":715,"__hash__":716},"courses/courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/02-platform-as-a-service.md",{"type":20,"value":518,"toc":657},[519,523,526,528,531,534,538,593,596,600,625,629,635,638,642,645,649,652,654],[23,520,522],{"id":521},"what-if-you-never-wanted-to-touch-the-operating-system-at-all","What if you never wanted to touch the operating system at all",[28,524,525],{},"The previous lesson ended with you owning an EC2 instance: patching its OS, installing a runtime, configuring a firewall. Now picture a solo developer with a web app to ship by Friday and zero interest in any of that. They do not want to choose an AMI, size a load balancer, or read a patch changelog. They want to hand over code and have it run. Platform as a Service is the model built for exactly that handoff.",[23,527,34],{"id":33},[28,529,530],{},"NIST defines PaaS as the capability to \"deploy onto the cloud infrastructure consumer-created or acquired applications created using programming languages, libraries, services, and tools supported by the provider.\" The consumer does not manage the underlying infrastructure, including network, servers, operating systems, or storage, but does control the deployed application and, often, configuration settings for the environment that hosts it.",[28,532,533],{},"Read that against the IaaS definition from the last lesson and the shift is precise: the operating system, which was yours in IaaS, now belongs to the provider. What is left for you is the application itself and the settings around how it runs.",[23,535,537],{"id":536},"who-manages-what-updated","Who manages what, updated",[46,539,540,550],{},[49,541,542],{},[52,543,544,546,548],{},[55,545,57],{},[55,547,207],{},[55,549,208],{},[62,551,552,561,569,577,585],{},[52,553,554,557,559],{},[67,555,556],{},"Networking, servers, virtualization",[67,558,72],{},[67,560,72],{},[52,562,563,565,567],{},[67,564,91],{},[67,566,94],{},[67,568,72],{},[52,570,571,573,575],{},[67,572,99],{},[67,574,94],{},[67,576,72],{},[52,578,579,581,583],{},[67,580,106],{},[67,582,94],{},[67,584,94],{},[52,586,587,589,591],{},[67,588,113],{},[67,590,94],{},[67,592,94],{},[28,594,595],{},"The whole shift is in one row: the operating system and runtime cross the line from \"you\" to \"provider.\" That is the entire difference between the 2 models, and it deserves more attention than a single row suggests, because it is exactly the line an exam scenario tests.",[23,597,599],{"id":598},"real-products-that-deliver-paas","Real products that deliver PaaS",[132,601,602,608,619],{},[135,603,604,607],{},[138,605,606],{},"AWS Elastic Beanstalk:"," packages EC2 instances, a load balancer, and auto-scaling behind one deployment, still built on AWS's IaaS underneath, but you never touch those pieces directly",[135,609,610,613,614,618],{},[138,611,612],{},"Heroku:"," you run ",[615,616,617],"code",{},"git push heroku main",", and Heroku detects your language, installs dependencies, and deploys, with zero server configuration for standard frameworks",[135,620,621,624],{},[138,622,623],{},"Google App Engine and Azure App Service:"," the same trade, upload code, the platform runs it",[23,626,628],{"id":627},"worked-example-the-same-app-2-ways","Worked example: the same app, 2 ways",[28,630,631,632,634],{},"Shipping a small Ruby web app through EC2 means launching an instance, installing Ruby and its dependencies, configuring a web server, attaching a load balancer, and setting up an auto-scaling group, several separate steps, each one yours to get right and keep patched. Shipping the same app through Heroku means running ",[615,633,617],{},". Heroku reads the code, recognizes it needs a Ruby runtime, builds it, and starts serving traffic, all from that 1 command.",[28,636,637],{},"That is not a smaller version of the same job. It is a different job: you stopped managing infrastructure and started managing only your application.",[23,639,641],{"id":640},"the-misconception-paas-is-not-zero-infrastructure-decisions","The misconception: PaaS is not \"zero infrastructure decisions\"",[28,643,644],{},"It is tempting to assume PaaS removes infrastructure thinking completely. It does not. You still pick a dyno or instance size, still set environment variables, and still pay for the capacity you have allocated whether or not it is handling traffic right now, a Heroku dyno you provision keeps running, and keeps billing, until you scale it down yourself. A model that only charges for the moment your code actually executes is a different idea again, serverless computing, and this course covers it later in the Modern Cloud Architectures topic. PaaS narrows your infrastructure job. It does not delete it.",[23,646,648],{"id":647},"the-boundary-ahead-paas-versus-saas","The boundary ahead: PaaS versus SaaS",[28,650,651],{},"PaaS still asks you to write the application. That is the line the next lesson erases entirely.",[23,653,177],{"id":176},[28,655,656],{},"PaaS trades some of the control IaaS gave you for speed: you stop managing servers and start shipping code straight to a platform that runs it. The next lesson covers what happens when even the code is no longer yours to write.",{"title":182,"searchDepth":183,"depth":183,"links":658},[659,660,661,662,663,664,665,666],{"id":521,"depth":186,"text":522},{"id":33,"depth":186,"text":34},{"id":536,"depth":186,"text":537},{"id":598,"depth":186,"text":599},{"id":627,"depth":186,"text":628},{"id":640,"depth":186,"text":641},{"id":647,"depth":186,"text":648},{"id":176,"depth":186,"text":177},[],[511,512,513,514],{},"/courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/02-platform-as-a-service",{"passingScore":199,"questions":672},[673,679,683,690,698,706],{"question":674,"type":203,"options":675,"correctAnswer":208,"explanation":678},"A developer wants to deploy a web app without choosing an operating system, patching it, or configuring a web server manually. Which service model fits?",[207,676,208,677],"Traditional on-premises hosting","None of these","PaaS takes the operating system and runtime off the developer's plate, leaving only the application code and its configuration. IaaS still requires choosing and patching an OS, and traditional on-premises hosting requires even more of that work.",{"question":680,"type":203,"options":681,"correctAnswer":214,"explanation":682},"In Platform as a Service, the customer is still responsible for patching the operating system the application runs on.",[213,214],"PaaS is defined by the provider taking over the operating system and runtime, which is exactly what separates it from IaaS. The customer keeps control of the application code and its configuration, but OS patching moves to the provider.",{"question":684,"type":203,"options":685,"correctAnswer":687,"explanation":689},"Compared to IaaS, what does PaaS additionally take off the customer's plate?",[219,686,687,688],"The physical network","The operating system and runtime","Application code and data","The single row that moves between IaaS and PaaS is the operating system and runtime, both of which shift from the customer to the provider. Application code and data stay with the customer in both models.",{"question":691,"type":226,"options":692,"correctAnswers":696,"explanation":697},"Which of the following are PaaS products? (Select all that apply.)",[693,240,694,695],"AWS EC2","Heroku","Google App Engine",[240,694,695],"Elastic Beanstalk, Heroku, and Google App Engine all let you deploy code without managing the operating system underneath it, the defining PaaS trait. AWS EC2 hands you a virtual machine with the OS layer still yours to manage, which makes it IaaS instead.",{"question":699,"type":203,"options":700,"correctAnswer":704,"explanation":705},"A team deploys their app on Heroku and assumes they no longer pay for anything unless their app is actively serving a request. Why is this assumption wrong?",[701,702,703,704],"Heroku does not charge for dynos at all","Heroku is a SaaS product, not a PaaS product","PaaS only charges for storage, never for compute","A provisioned dyno keeps running, and billing, until you scale it down yourself, whether or not it is handling traffic","PaaS still requires choosing and paying for allocated capacity, a dyno or instance size, that keeps running and billing regardless of whether it is currently serving traffic. A model that only charges for the moment code actually executes is a different idea, serverless computing, covered later in this course.",{"question":707,"type":203,"options":708,"correctAnswer":709,"explanation":713},"Deploying an app via Heroku instead of directly via EC2 mainly changes which job the developer is doing. What is that change?",[709,710,711,712],"The developer stops managing infrastructure and manages only the application","The developer now manages the physical hardware instead of AWS","The developer takes on more responsibility for network security","There is no real change, the number of steps stays the same","Shipping through EC2 means installing a runtime, configuring a web server, and setting up a load balancer and auto-scaling group, several separate steps the developer owns. Shipping through Heroku collapses that into a single deploy command, because the developer's job has shifted from managing infrastructure to managing only the application.",{"title":508,"description":509},"courses/cloud-computing-fundamentals/en/domains/02-cloud-services-and-architecture/01-cloud-service-models/02-platform-as-a-service","ERzC9aKGOJ1IGuITkecV6OlXj68W2tYhsKzklnJiklM"]