section 6 Using Virtualization and Databases

What’s the difference between Virtual machines and containers?

Virtual machines rely on a layer that virtualizes hardware, whereas containers rely on a virtualized operating system

Virtualization

It enables the abstraction of physical hardware, allowing multiple virtual machines (VMs) or containers to run on a single physical server, optimizing resource utilization, scalability, and cost efficiency

Virtualization services are provided by a component called a what?

Hypervisor

Hypervisor

This layer manages the hardware allocations and controls VM access to hardware

. VMs can be restored to a given point in time with what?

Snapshots

Deploy Virtual Machines in the Cloud

You'll need to name the VM and select its hardware settings. Also need to define cloud-specific values, such as regions, authentication options, metering identifiers, and other choices

Configuration files

Define virtual machine (VM) parameters that specify all the hardware allocations—Central Processing Unit (CPU) information, random-access memory (RAM) quantities, network options, and storage—for the VM

Can be used as templates to construct identical VMs with preconfigured settings

Clusters

Are groups of computers configured to work together as a single unit. Each cluster member is called a node

Provide fault tolerance. If one node of the cluster fails, the other systems can maintain services

What does Clusters rely on?

Isolated management network that allows the nodes to communicate with each other to maintain status and add or remove nodes

Clients access cluster services via a separate network

Host Affinity

Associates or links one or more cluster node VMs with cluster host systems.

Affinity rules typically attempt to keep VM nodes on the same host to enhance performance (node communications occur using the host bus rather than the network)

Anti-affinity

In virtualization, a deliberate distribution of virtual machine instances across multiple hosts to avoid single points of failure.

(if the sole host of five VMs in a cluster fails, all five VMs become unavailable)

Hardware Pass-Through

Virtualization feature that allows a virtual machine (VM) to directly access physical hardware components, such as GPUs

Cloning

The process of quickly duplicating a virtual machine's configuration when several identical machines are needed immediately.

Virtual Machine Cloning

Provide a snapshot of their current configuration and any data stored on the VM

Overlay Networks

Abstract media, switches, and routers to enable administrators to define logical networks without worrying about the physical network infrastructure

Aids in IP address configurations (without consuming production IP addresses), load balancing, and Network Address Translation (NAT) services for external communications

The benefits of overlay networks include:

Simplified administration

Greater flexibility

Greater scalability

Increased security through better segmentation

Improved efficiency

Virtual Extensible LAN (VXLAN)

Technology used to implement an overlay network so that hosts in separate subnets can establish layer 2 adjacency in a discrete logical segment. The 24-bit VXLAN ID space supports up to 16 million logical segments.

Generic Routing Encapsulation (GRE)

Tunneling protocol allowing the transmission of encapsulated frames or packets from different types of network protocol over an IP network.

Virtual Machine Networks

Allow VMs to communicate like physical devices on a network using virtual switches and configurable network settings

Virtual Machine Network Types: External

Allow access to the host's network interface card and the physical network beyond it.

Virtual Machine Network Types: Internal

Allow network access to the host itself (but not its network card) and other VMs on the host but no external network access.

Virtual Machine Network Types: Private

Allow network access only to other VMs but not to the host or the network beyond it

Local Storage

Easy to install and configure but very difficult to scale and difficult to share directly with other services

Takes the form of solid-state drives (SSDs) or hard disk drives (HDDs)

Network Attached Storage (NAS)

A dedicated storage device connected to a network, allowing multiple devices and users to access shared files over standard network protocols (e.g., NFS, SMB/CIFS)

Provides centralized, scalable, and remote-accessible storage, commonly used for file sharing, backups, and media streaming

Storage Area Network (SAN)

Devices provide greater scalability, fault tolerance, and performance than do NAS devices. They, however, are also significantly more complex and more expensive

Solutions may be found in private cloud data centers

What are the three primary components that make up a complete SAN solution?

The first component is one or more servers that manage access to the data. The second component is an isolated network between the servers and the storage infrastructure. The final component is the storage infrastructure itself

Adding a NAS to an already overwhelmed network can cause what?

Increases network traffic and may result in unacceptable delays for users and applications to access data

First component of SAN

The servers may be Microsoft Windows Server installations. The server hardware will likely contain Host Bus Adapter (HBA) cards for access to the storage infrastructure

Second component of SAN

The isolated network is the communications path between the servers and the storage infrastructure. There is typically a great deal of redundancy in the network to ensure that data access is highly available

Third component of SAN

The storage infrastructure includes the actual storage drives which are housed in separate boxes from the servers. The disk arrays consist of many HDDs, controllers, and supporting components to connect to the SAN network

Containers

Complete, portable solution. It contains the application code, runtime, libraries, settings, and other components—everything needed for the software to run.

This complete package is portable and will run on any platform hosting a container engine, including the cloud infrastructure

Share a single operating system (OS) (usually Linux), and provide a single function

What do Container engines use for container image formats?

Open Container Initiative (OCI) standard

What is the first step in deploying containers on a local system?

Selecting and installing a container engine on a server

What are the three components to a container solution?

Configuration file

Container image

Running container

Configuration file

Think of this file as a specifications sheet or a set of instructions

container image

The image is a template for how containers should look when running. The image is built from the instructions provided in the Dockerfile

Last, containers are run

Containers are executing instances of an image. Many identical containers can be spawned from a single image

Containers cannot run without what?

An image defining their configuration, and images must be built from the configuration file

Containers are stateless, which means what?

Any changes to the container or applications running in it are not conserved. Persistent changes to the container occur via configuration file updates

It's easy to begin working with containers on a local system, how is it done?

Select and install a container engine (Docker is a good place to start) on a workstation or server.

Next, pull one or more images from a container registry. Images usually offer specific capabilities, so find one you can work with

Container image registries

Are storage repositories for the container images your applications and developers need. These registries offer a place to store, secure, and share images

You can integrate them into your DevOps and Continuous Integration/Continuous Deployment (CI/CD) processes. That means orchestration processes that automate a complete service build can pull images from these registries on the fly.

Private registries

Make it easier to maintain images and secure resources.

May be on-premises or in a remote data center (private cloud)

Cloud service providers offer many container services that integrate with your organization's needs. Here are a few examples:

Amazon Elastic Container Service: Run, monitor, and scale apps with integration with other Amazon Web Services (AWS) services.

Azure Kubernetes Service: Run and scale containers on the Kubernetes platform.

Google Compute Engine: Run and manage containers on Kubernetes, Docker, and other platforms.

For networking the default configuration on a container is what?

Only allows containers to communicate with other containers on the internal container network. The container cannot see or access services on the host or an external network

Port Mapping

Associates container ports with host ports, providing an access path between the container and outside network functionality

Ex. When deploying an Nginx container, the web service is only accessible to other containers by default. The host system and external clients cannot access

Ephemeral storage

Does not persist between container restarts or crashes. Instead, it is created and removed along with the container. It's used for applications or services that don't need to store permanent data.

Ex: Data caching, Container logging, Miscellaneous scratch space

Persistent storage

Is independent of the container's lifecycle. It stores data even if the container is stopped or deleted. This form of storage is necessary for stateful applications that generate data that must exist after the container itself is removed

Container clusters share the persistent storage space, providing access to the data to all nodes.

Orchestration

Enables many options for organizations using DevOps approaches.

In general provides a sequential management chain spanning multiple automated tasks.

It allows organizations to deploy and manage huge quantities of apps and their related microservices using containers

Relational Database

Uses tables of columns and rows to organize structured data

Queried by using the Structured Query Language (SQL) and may be referred to as "SQL databases."

Tables consist of columns and rows.

Column: Data attributes

Row: Instances of that data

Non-relational Database

Uses various specialized designs to support different unstructured data types

Suppose your organization wants to store information about its products retrieved from customer social media posts. The data might include user profiles, posts, likes, comments, and more. Some data might be images, video, or even audio files

Structured Query Language (SQL)

Programming and query language common to many relational database management systems.

NoSQL

Designed to hold unstructured information organized by type. This design leads to a great deal of flexibility, application scalability, and rapid data growth. This makes them a perfect candidate for cloud hosting.

Four common non-relational databases are:

Document-oriented, Key-value, Wide column, Graph stores

Self-managed databases

Are handled by your internal staff and hosted on your business platform (whether on-premises or in a private cloud)

Require your attention for design, deployment, management, maintenance, and scaling. Database administrators are responsible for all aspects of the databases

Provider-managed databases

Are hosted in a public cloud. Various factors govern your organization's choice. In some cases, different company databases will be hosted in different environments

Self-managed advantages:

• Control of data

• Flexibility

• Avoids vendor lock-in

• Control over database choices

Self-managed disadvantages:

• Significant investment in time and resources

• Technical skills

• Commitment

Provider-managed advantages:

Structured

Maintenance-free/support provided

Automation

Probable cost savings and pay-as-you-go cost model

Built-in scalability

Provider-managed disadvantages:

Possible vendor lock-in

Security concerns in public clouds

Fewer customization options

Data sovereignty issues

Direct migrations

Moving database contents between two providers (or your private cloud) within the same database design (such as MS SQL Server to MS SQL Server)

Cross-service migrations

Data conversions between two different database engines (such as Oracle to Amazon Aurora)

Considering the need for persistent data storage in a containerized application, which approach would be most effective?

Implement external storage solutions like volumes or bind mounts.

Ensuring data is preserved across container restarts

Your team needs to deploy a container instance in Azure for a demo application.

Which steps should you follow to ensure the container is properly set up, based on the document?

Select to create container instance, name the container, select an image source registry, and configure network settings.

A financial services company is considering moving their on-premises database to a cloud-hosted solution. They need to ensure high availability and scalability while maintaining control over their data.

What steps should they take?

Build and manage their own database structure on cloud Infrastructure as a Service (IaaS) solutions.

Primary keys

Ensure no data in a row is duplicated, making each row in the table unique

A network technician is diagnosing connections to their companies' Amazon Web Services (AWS) sites.

What should the technician check?

Log files

ISP outage

CSP outage