Chapter 2: Overview of Cloud Computing and Google Cloud Flashcards
In this chapter, we’ll cover: Cloud computing basics How Google Cloud differs from other cloud platforms The various Google Cloud products and services of relevance to the cloud architect The various ways you can access and manage your cloud The business and technical context of cloud architecture
Public Cloud
A public cloud offers cloud services to the public; think of the big providers such as Google Cloud, Amazon Web Services (AWS), Microsoft Azure, and Alibaba Cloud.
Private Cloud
A private cloud is developed in-house, specific to a business. Think of the Facebook Cloud, which Facebook had built for its own use. Or think of an on-premises cloud, where all servers, storage, and networks are dedicated to the company and hosted in a dedicated data center/s.
Hybrid Cloud
A hybrid cloud uses a mixture of a public cloud and a private cloud to create a more diverse environment. Customers with highly sensitive data, such as financial services or healthcare, like to have their own environments under their full control, where they can store their most highly sensitive workloads and data. Or maybe the customer hasn’t gotten around to migrating fully to a public cloud because of a lack of commitment, features, time, and so on.
Multi-Cloud
A multi-cloud is similar to a hybrid cloud, in that multiple clouds are at play, but typically in a multi-cloud environment, a business will be using multiple public clouds. Many top companies today use this model, such as Snapchat, which uses both Google Cloud and AWS for its workloads.
- You are looking to buy a new computer for personal use. You want a powerful enough computer to surf websites and run (mostly gaming) applications. But you realize that you are short on cash. What computer should you buy?
- MacBook Pro
- Chromebook
- Windows 10 desktop
- Windows 10 laptop
- You are looking to buy a new computer for personal use. You want a powerful enough computer to surf websites and run (mostly gaming) applications. But you realize that you are short on cash. What computer should you buy?
* Windows 10 desktop
This question is about parsing requirements. First, the requirement of surfing websites can be accommodated by all the computers listed. Second, running locally installed applications, including games, eliminates Chromebook from the picture and even the MacBook. Third, you have a requirement for the most performance with the least cost. Lastly, while both the Windows desktop and laptop seem to be strong answers, the prerequisite knowledge of the gaming PC space will help you answer this question effectively, making the Windows desktop the best answer, because these computers typically offer the most bang for the buck. You’ll see questions on the exam that are very tricky, just like this one.
- Your company has made plans to roll out OpenShift, a Kubernetes platform solution offered by IBM Red Hat, across all its on-premises and public cloud environments. Given that you are the lead architect responsible for your company’s GCP deployments, what type of shared responsibility model will this deployment entail for you?
- On-premises
- IaaS
- PaaS
- SaaS
- Your company has made plans to roll out OpenShift, a Kubernetes platform solution offered by IBM Red Hat, across all its on-premises and public cloud environments. Given that you are the lead architect responsible for your company’s GCP deployments, what type of shared responsibility model will this deployment entail for you?
* IaaS
The key to remember here is that for a service provided (GCP in this case) to take responsibility for its PaaS, it must offer the service as a managed service. GCP offers its own Kubernetes platform called GKE. But OpenShift is not a Google-offered PaaS solution. As such, Google will not take responsibility for the backend operations and design of your OpenShift environments. You will need to manage all the VMs that OpenShift will provision as part of its GCP deployment. So this is an IaaS deployment from a shared responsibility model perspective.
- VPC networks are:
- Global
- Regional
- Zonal
- Local
- VPC networks are:
* Global
VPC networks are global in GCP. Most questions on the exam are multilayered. So you will generally not get a straight question like this. Knowing this element will help you answer multilayered questions more effectively.
- Subnets are:
- Global
- Regional
- Zonal
- Local
- Subnets are:
* Regional
Subnets are regional within GCP. That means that you can create instances within your subnet across multiple zones. Keep that in mind when building highly available solutions. Many other cloud service providers support only zonal subnets.
- You need to attach high-performance storage with very high IOPS and low latency to your VM instance. Which technology should you use?
- Google Cloud Storage
- Local SSD
- Cloud FileStore
- Persistent SSD disk
- You need to attach high-performance storage with very high IOPS and low latency to your VM instance. Which technology should you use?
* Local SSD
From a storage perspective, you can attach each of the provided answers to your VMs. Google Cloud Storage can be mounted to a VM. But this is not the highest performer. The highest performance possible is achieved when the storage is included with the CPU and not attached via the network. The only answer that satisfies that criteria is Local SSD. Its read and write IOPS are almost ten times higher than persistent network attached storage options.
- Google’s operational excellence principle demands the building of a foundation to enable reliability successfully across your infrastructure by efficiently running, managing, and monitoring systems that deliver business value. Which of the following is not a key strategy that drives this principle?
- End-to-end automation
- Monitoring business objectives
- Performance and cost optimization
- Disaster recovery
- Google’s operational excellence principle demands the building of a foundation to enable reliability successfully across your infrastructure by efficiently running, managing, and monitoring systems that deliver business value. Which of the following is not a key strategy that drives this principle?
* Performance and cost optimization
Operational excellence does not include performance and cost optimization; that is a different system design principle.
- Which of the following service level measures are considered a legally enforceable contract between the service provider and the service consumer?
- SLA
- SLE
- SLO
- SLI
- Which of the following service level measures are considered a legally enforceable contract between the service provider and the service consumer?
* SLA
A service level agreement (SLA) is an enforceable legal contract between a service provider and a service customer. A service level indicator (SLI) is a measure of the service level provided by a service provider to a customer. SLIs form the basis of service level objectives (SLOs), which in turn form the basis of SLAs. In this way, SLAs define the level of service expected by a customer from a supplier, laying out the metrics by which that service is measured and the remedies or penalties, if any, should the agreed-upon service levels not be achieved.
- Your development team is building a new business-critical application using virtual machines to be deployed in a dedicated production project. As part of this effort, the team is looking to implement a dedicated application testing environment within a development project. The tests generally take less than an hour to complete. They need to keep the testing machine costs low, but consistent with their production environment. Which type of virtual machine optimization strategy would you use for their testing project environment?
- Use sole-tenant nodes.
- Automate the VM life cycle.
- Use preemptible VMs.
- Use purchase commitments.
- Your development team is building a new business-critical application using virtual machines to be deployed in a dedicated production project. As part of this effort, the team is looking to implement a dedicated application testing environment within a development project. The tests generally take less than an hour to complete. They need to keep the testing machine costs low, but consistent with their production environment. Which type of virtual machine optimization strategy would you use for their testing project environment?
* Automate the VM life cycle.
This is typically the type of question you might see on the exam. You’re provided a set of requirements that can often be met by a number of approaches. Between B and C, there are many cost savings. In fact, I would try to use automated preemptible VMs to get the most bang for my dollars. Preemptible machines give you up to 80 percent off and would satisfy the ability to use the same type of machines as the production systems (as do purchase commitments). Furthermore, with a one-hour timetable to run your tests, preemptible VMs are ideal. Even if you get preempted, you can try again later. If you automate your VM life cycle, you immediately get about a 96 percent savings. That means you run the machine for only 1 hour of 24 hours a day. The ideal answer in real life could be to use both preemptible instances and VM life cycle management as part of an automated CICD build and test process. But that is not how the question presented things. So the biggest contributor to cost savings would be automation of the VM life cycle.
- Which Google Cloud Platform database offering is best suited for integration with client-side mobile and web applications, gaming leaderboards, and user presence at global scale?
- BigQuery
- Cloud Memorystore
- Cloud Bigtable
- Cloud Firestore
- Which Google Cloud Platform database offering is best suited for integration with client-side mobile and web applications, gaming leaderboards, and user presence at global scale?
* Cloud Firestore
You can argue that certain database technologies are cross-purpose and could be deployed in other situations. You would be right. But, again, the key to passing the test is to provide the best answer based on the key information presented. Firestore would be the best database to integrate with mobile and web apps.
- You decide to use GCP to host a simple website using Drupal Content Management from Google Cloud Platform Marketplace to run a Google Compute Engine VM instance. You have global ambitions but a limited budget. What feature would you enable to provide a better experience to your global audience?
- Cloud Interconnect
- Cloud DNS
- Cloud CDN
- Cloud Load Balancer
- You decide to use GCP to host a simple website using Drupal Content Management from Google Cloud Platform Marketplace to run a Google Compute Engine VM instance. You have global ambitions but a limited budget. What feature would you enable to provide a better experience to your global audience?
* Cloud CDN
The only answer that helps with the content distribution is Cloud CDN (Content Delivery Network). Although DNS may be in use, it won’t make an impact on the global audience, which this question is asking for. Furthermore, since this is a simple VM, adding a load balancer to the architecture does not buy you anything. To leverage a load balancer, you would want multiple VMs within multiple regions, which is clearly not the case in this question.
