AWS Compute - Part 4: Load Balancer and Autoscaling

Hulk Pham - Oct 31 - - Dev Community

TL;DR

High Availability and Load Balancing

  • High availability is crucial for systems, often expressed as a percentage of uptime or number of nines
  • Elastic Load Balancing (ELB) distributes incoming traffic across multiple targets, improving availability and scalability
  • ELB offers three types: Application Load Balancer (ALB), Network Load Balancer (NLB), and Gateway Load Balancer (GLB), each suited for different use cases

Amazon EC2 Auto Scaling

  • EC2 Auto Scaling automatically adds or removes EC2 instances based on defined policies, ensuring optimal performance and cost-efficiency
  • Auto Scaling groups define where resources are deployed, specifying VPC, subnets, and instance purchase options
  • Launch templates or configurations specify the resources to be scaled, including AMI, instance type, and security groups
  • Scaling policies determine when to add or remove instances, using CloudWatch metrics and alarms to trigger actions

I. High Availability

The availability of a system is typically expressed as a percentage of uptime in a given year or as a number of nines. In the following table is a list of availability percentages based on the downtime per year and its notation in nines.

Availability (%) Downtime (per year)
90% (one nine of availability) 36.53 days
99% (two nines of availability) 3.65 days
99.9% (three nines of availability) 8.77 hours
99.95% (three and a half nines of availability) 4.38 hours
99.99% (four nines of availability) 52.60 minutes
99.995% (four and a half nines of availability) 26.30 minutes
99.999% (five nines of availability) 5.26 minutes

To increase availability, you need redundancy. This typically means more infrastructure—more data centers, more servers, more databases, and more replication of data. You can imagine that adding more of this infrastructure means a higher cost. Customers want the application to always be available, but you need to draw a line where adding redundancy is no longer viable in terms of revenue.

1. Why improve application availability?

In the current application, one EC2 instance hosts the application. The photos are served from Amazon S3, and the structured data is stored in Amazon DynamoDB. That single EC2 instance is a single point of failure for the application.

Even if the database and Amazon S3 are highly available, customers have no way to connect if the single instance becomes unavailable. One way to solve this single point of failure issue is to add one more server in a second Availability Zone.

2. Adding a second Availability Zone

The physical location of a server is important. In addition to potential software issues at the operating system (OS) or application level, you must also consider hardware issues. They might be in the physical server, the rack, the data center, or even the Availability Zone hosting the virtual machine. To remedy the physical location issue, you can deploy a second EC2 instance in a second Availability Zone. This second instance might also solve issues with the OS and the application.

VPC_subnets description

However, when there is more than one instance, it brings new challenges, such as the following:

  • Replication process – The first challenge with multiple EC2 instances is that you need to create a process to replicate the configuration files, software patches, and application across instances. The best method is to automate where you can.
  • Customer redirection – The second challenge is how to notify the clients—the computers sending requests to your server—about the different servers. You can use various tools here. The most common is using a Domain Name System (DNS) where the client uses one record that points to the IP address of all available servers.

    However, this method isn't always used because of propagation — the time frame it takes for DNS changes to be updated across the Internet.

    Another option is to use a load balancer, which takes care of health checks and distributing the load across each server. Situated between the client and the server, a load balancer avoids propagation time issues. You will learn more about load balancers in the next section.

  • Types of high availability – The last challenge to address when there is more than one server is the type of availability you need: active-passive or active-active.

3. High availability categories.

Active-passive systems

With an active-passive system, only one of the two instances is available at a time. One advantage of this method is that for stateful applications (where data about the client’s session is stored on the server), there won’t be any issues. This is because the customers are always sent to the server where their session is stored.

Active-active systems

A disadvantage of an active-passive system is scalability. This is where an active-active system shines. With both servers available, the second server can take some load for the application, and the entire system can take more load. However, if the application is stateful, there would be an issue if the customer’s session isn’t available on both servers. Stateless applications work better for active-active systems.

II. Elastic Load Balancing

The Elastic Load Balancing (ELB) service can distribute incoming application traffic across EC2 instances, containers, IP addresses, and Lambda functions.

1. Load balancers

Load balancing refers to the process of distributing tasks across a set of resources. In the case of the Employee Directory application, the resources are EC2 instances that host the application, and the tasks are the requests being sent. You can use a load balancer to distribute the requests across all the servers hosting the application.

To do this, the load balancer needs to take all the traffic and redirect it to the backend servers based on an algorithm. The most popular algorithm is round robin, which sends the traffic to each server one after the other.

A typical request for an application starts from a client's browser. The request is sent to a load balancer. Then, it’s sent to one of the EC2 instances that hosts the application. The return traffic goes back through the load balancer and back to the client's browser.

Although it is possible to install your own software load balancing solution on EC2 instances, AWS provides the ELB service for you.

ELB_arch description

2. ELB features

The ELB service provides a major advantage over using your own solution to do load balancing. Mainly, you don’t need to manage or operate ELB. It can distribute incoming application traffic across EC2 instances, containers, IP addresses, and Lambda functions. Other key features include the following:

  • Hybrid mode – Because ELB can load balance to IP addresses, it can work in a hybrid mode, which means it also load balances to on-premises servers.
  • High availability – ELB is highly available. The only option you must ensure is that the load balancer's targets are deployed across multiple Availability Zones.
  • Scalability – In terms of scalability, ELB automatically scales to meet the demand of the incoming traffic. It handles the incoming traffic and sends it to your backend application.

3. Health checks

Monitoring is an important part of load balancers because they should route traffic to only healthy EC2 instances. That’s why ELB supports two types of health checks as follows:

  • Establishing a connection to a backend EC2 instance using TCP and marking the instance as available if the connection is successful.
  • Making an HTTP or HTTPS request to a webpage that you specify and validating that an HTTP response code is returned.

Taking time to define an appropriate health check is critical. Only verifying that the port of an application is open doesn’t mean that the application is working. It also doesn’t mean that making a call to the home page of an application is the right way either.

For example, the Employee Directory application depends on a database and Amazon S3. The health check should validate all the elements. One way to do that is to create a monitoring webpage, such as /monitor. It will make a call to the database to ensure that it can connect, get data, and make a call to Amazon S3. Then, you point the health check on the load balancer to the /monitor page.

health_check description

After determining the availability of a new EC2 instance, the load balancer starts sending traffic to it. If ELB determines that an EC2 instance is no longer working, it stops sending traffic to it and informs Amazon EC2 Auto Scaling. It is the responsibility of Amazon EC2 Auto Scaling to remove that instance from the group and replace it with a new EC2 instance. Traffic is only sent to the new instance if it passes the health check.

If Amazon EC2 Auto Scaling has a scaling policy that calls for a scale down action, it informs ELB that the EC2 instance will be terminated. ELB can prevent Amazon EC2 Auto Scaling from terminating an EC2 instance until all connections to the instance end. It also prevents any new connections. This feature is called connection draining. We will learn more about Amazon EC2 Auto Scaling in the next lesson.

4. ELB components

The ELB service is made up of three main components: rules, listeners, and target groups.

ELB_service_updated description

Rule

To associate a target group to a listener, you must use a rule. Rules are made up of two conditions. The first condition is the source IP address of the client. The second condition decides which target group to send the traffic to.

Listener

The client connects to the listener. This is often called client side. To define a listener, a port must be provided in addition to the protocol, depending on the load balancer type. There can be many listeners for a single load balancer.

Target group

The backend servers, or server side, are defined in one or more target groups. This is where you define the type of backend you want to direct traffic to, such as EC2 instances, Lambda functions, or IP addresses. Also, a health check must be defined for each target group.

5. Types of load balancers

We will cover three types of load balancers: Application Load Balancer (ALB), Network Load Balancer (NLB), and Gateway Load Balancer (GLB).

A. Application Load Balancer

For our Employee Directory application, we are using an Application Load Balancer. An Application Load Balancer functions at Layer 7 of the Open Systems Interconnection (OSI) model. It is ideal for load balancing HTTP and HTTPS traffic. After the load balancer receives a request, it evaluates the listener rules in priority order to determine which rule to apply. It then routes traffic to targets based on the request content.

Primary features of an Application Load Balancer:

  • Routes traffic based on request data

    An Application Load Balancer makes routing decisions based on the HTTP and HTTPS protocol. For example, the ALB could use the URL path (/upload) and host, HTTP headers and method, or the source IP address of the client. This facilitates granular routing to target groups.

  • Sends responses directly to the client

    An Application Load Balancer can reply directly to the client with a fixed response, such as a custom HTML page. It can also send a redirect to the client. This is useful when you must redirect to a specific website or redirect a request from HTTP to HTTPS. It removes that work from your backend servers.

  • Uses TLS offloading

    An Application Load Balancer understands HTTPS traffic. To pass HTTPS traffic through an Application Load Balancer, an SSL certificate is provided one of the following ways:

    • Importing a certificate by way of IAM or ACM services
    • Creating a certificate for free using ACM

    This ensures that the traffic between the client and Application Load Balancer is encrypted.

  • Authenticates users

    An Application Load Balancer can authenticate users before they can pass through the load balancer. The Application Load Balancer uses the OpenID Connect (OIDC) protocol and integrates with other AWS services to support protocols, such as the following:

    • SAML
    • Lightweight Directory Access Protocol (LDAP)
    • Microsoft Active Directory
    • Others
  • Secures traffic

    To prevent traffic from reaching the load balancer, you configure a security group to specify the supported IP address ranges.

  • Supports sticky sessions

    If requests must be sent to the same backend server because the application is stateful, use the sticky session feature. This feature uses an HTTP cookie to remember which server to send the traffic to across connections.

B. Network Load Balancer

A Network Load Balancer is ideal for load balancing TCP and UDP traffic. It functions at Layer 4 of the OSI model, routing connections from a target in the target group based on IP protocol data.

Primary features of an Network Load Balancer:

  • Sticky sessions Routes requests from the same client to the same target.
  • Low latency Offers low latency for latency-sensitive applications.
  • Source IP address Preserves the client-side source IP address.
  • Static IP support Automatically provides a static IP address per Availability Zone (subnet).
  • Elastic IP address support Lets users assign a custom, fixed IP address per Availability Zone (subnet).
  • DNS failover Uses Amazon Route 53 to direct traffic to load balancer nodes in other zones.

C. Gateway Load Balancer

A Gateway Load Balancer helps you to deploy, scale, and manage your third-party appliances, such as firewalls, intrusion detection and prevention systems, and deep packet inspection systems. It provides a gateway for distributing traffic across multiple virtual appliances while scaling them up and down based on demand.

Primary features of an Gateway Load Balancer:

  • High availability Ensures high availability and reliability by routing traffic through healthy virtual appliances.
  • Monitoring Can be monitored using CloudWatch metrics.
  • Streamlined deployments Can deploy a new virtual appliance by selecting it in the AWS Marketplace.
  • Private connectivity

    Connects internet gateways, virtual private clouds (VPCs), and other network resources over a private network.

6. Selecting between ELB types

You can select between the ELB service types by determining which feature is required for your application. The following table presents a list of some of the major features of load balancers. For a complete list, see "Elastic Load Balancing features" in the Resources section at the end of this lesson.

Feature ALB NLB GLB
Load Balancer Type Layer 7 Layer 4 Layer 3 gateway and Layer 4 load balancing
Target Type IP, instance, Lambda IP, instance, ALB IP, instance
Protocol Listeners HTTP, HTTPS TCP, UDP, TLS IP
Static IP and Elastic IP Address Yes
Preserve Source IP Address Yes Yes Yes
Fixed Response Yes
User Authentication Yes

III. Amazon EC2 Auto Scaling

Amazon EC2 Auto Scaling helps you maintain application availability. You can automatically add or remove EC2 instances using scaling policies that you define.

1. Capacity issues

You can improve availability and reachability by adding one more server. However, the entire system can again become unavailable if there is a capacity issue. This section looks at load issues for both active-passive systems and active-active systems. These issues are addressed through scaling.

A. Vertical Scaling

Increase the instance size. If too many requests are sent to a single active-passive system, the active server will become unavailable and, hopefully, fail over to the passive server. But this doesn’t solve anything.

With active-passive systems, you need vertical scaling. This means increasing the size of the server. With EC2 instances, you select either a larger type or a different instance type. This can be done only while the instance is in a stopped state. In this scenario, the following steps occur:

  1. Stop the passive instance. This doesn’t impact the application because it’s not taking any traffic.
  2. Change the instance size or type, and then start the instance again.
  3. Shift the traffic to the passive instance, turning it active.
  4. Stop, change the size, and start the previous active instance because both instances should match.

When the number of requests reduces, you must do the same operation. Even though there aren’t that many steps involved, it’s actually a lot of manual work. Another disadvantage is that a server can only scale vertically up to a certain limit. When that limit is reached, the only option is to create another active-passive system and split the requests and functionalities across them. This can require massive application rewriting.

This is where the active-active system can help. When there are too many requests, you can scale this system horizontally by adding more servers.

B. Horizontal Scaling

Add additional instances. As mentioned, for the application to work in an active-active system, it’s already created as stateless, not storing any client sessions on the server. This means that having two or four servers wouldn’t require any application changes. It would only be a matter of creating more instances when required and shutting them down when traffic decreases. The Amazon EC2 Auto Scaling service can take care of that task by automatically creating and removing EC2 instances based on metrics from Amazon CloudWatch. We will learn more about this service in this lesson.

You can see that there are many more advantages to using an active-active system in comparison with an active-passive system. Modifying your application to become stateless provides scalability.

2. Traditional scaling compared to auto scaling

With a traditional approach to scaling, you buy and provision enough servers to handle traffic at its peak. However, this means at nighttime, for example, you might have more capacity than traffic, which means you’re wasting money. Turning off your servers at night or at times when the traffic is lower only saves on electricity.

The cloud works differently with a pay-as-you-go model. You must turn off the unused services, especially EC2 instances you pay for on-demand. You can manually add and remove servers at a predicted time. But with unusual spikes in traffic, this solution leads to a waste of resources with over-provisioning or a loss of customers because of under-provisioning.

The need here is for a tool that automatically adds and removes EC2 instances according to conditions you define. That’s exactly what the Amazon EC2 Auto Scaling service does.

3. Amazon EC2 Auto Scaling features

The Amazon EC2 Auto Scaling service adds and removes capacity to keep a steady and predictable performance at the lowest possible cost. By adjusting the capacity to exactly what your application uses, you only pay for what your application needs. This means Amazon EC2 Auto Scaling helps scale your infrastructure and ensure high availability.

Scaling features:

  • Automatic scaling Automatically scales in and out based on demand.Click to flip
  • Scheduled scaling Scales based on user-defined schedules.Click to flip
  • Fleet management Automatically replaces unhealthy EC2 instances.Click to flip
  • Predictive scaling Uses machine learning (ML) to help schedule the optimum number of EC2 instances.Click to flip
  • Purchase options Includes multiple purchase models, instance types, and Availability Zones.Click to flip
  • Amazon EC2 availability Comes with the Amazon EC2 service.Click to flip

4. ELB with Amazon EC2 Auto Scaling

Additionally, the ELB service integrates seamlessly with Amazon EC2 Auto Scaling. As soon as a new EC2 instance is added to or removed from the Amazon EC2 Auto Scaling group, ELB is notified. However, before ELB can send traffic to a new EC2 instance, it needs to validate that the application running on the EC2 instance is available.

This validation is done by way of the ELB health checks feature you learned about in the previous lesson.

IV. Configure Amazon EC2 Auto Scaling components

There are three main components of Amazon EC2 Auto Scaling. Each of these components addresses one main question as follows:

  • Launch template or configuration: Which resources should be automatically scaled?
  • Amazon EC2 Auto Scaling groups: Where should the resources be deployed?
  • Scaling policies: When should the resources be added or removed?

1. Launch templates and configurations

Multiple parameters are required to create EC2 instances—Amazon Machine Image (AMI) ID, instance type, security group, additional Amazon EBS volumes, and more. All this information is also required by Amazon EC2 Auto Scaling to create the EC2 instance on your behalf when there is a need to scale. This information is stored in a launch template.

You can use a launch template to manually launch an EC2 instance or for use with Amazon EC2 Auto Scaling. It also supports versioning, which can be used for quickly rolling back if there's an issue or a need to specify a default version of the template. This way, while iterating on a new version, other users can continue launching EC2 instances using the default version until you make the necessary changes.

A launch template specifies instance configuration information, such as the ID of the AMI, instance type, and security groups. You can have multiple versions of a launch template with a subset of the full parameters.

launch_template_2 description

You can create a launch template in one of three ways as follows:

  • Use an existing EC2 instance. All the settings are already defined.
  • Create one from an already existing template or a previous version of a launch template.
  • Create a template from scratch. These parameters will need to be defined: AMI ID, instance type, key pair, security group, storage, and resource tags.

Another way to define what Amazon EC2 Auto Scaling needs to scale is by using a launch configuration. It’s similar to the launch template, but you cannot use a previously created launch configuration as a template. You cannot create a launch configuration from an existing Amazon EC2 instance. For these reasons, and to ensure that you get the latest features from Amazon EC2, AWS recommends you use a launch template instead of a launch configuration.

2. Amazon EC2 Auto Scaling groups

The next component Amazon EC2 Auto Scaling needs is an Amazon EC2 Auto Scaling group. An Auto Scaling group helps you define where Amazon EC2 Auto Scaling deploys your resources. This is where you specify the Amazon Virtual Private Cloud (Amazon VPC) and subnets the EC2 instance should be launched in. Amazon EC2 Auto Scaling takes care of creating the EC2 instances across the subnets, so select at least two subnets that are across different Availability Zones.

With Auto Scaling groups, you can specify the type of purchase for the EC2 instances. You can use On-Demand Instances or Spot Instances. You can also use a combination of the two, which means you can take advantage of Spot Instances with minimal administrative overhead.

To specify how many instances Amazon EC2 Auto Scaling should launch, you have three capacity settings to configure for the group size.

To learn more, choose each of the three numbered markers.

auto_scaling_group description

Minimum capacity
This is the minimum number of instances running in your Auto Scaling group, even if the threshold for lowering the number of instances is reached.
When Amazon EC2 Auto Scaling removes EC2 instances because the traffic is minimal, it keeps removing EC2 instances until it reaches a minimum capacity. 
When reaching that limit, even if Amazon EC2 Auto Scaling is instructed to remove an instance, it does not. This ensures that the minimum is kept.
Note: Depending on your application, using a minimum of two is recommended to ensure high availability. However, you ultimately know how many EC2 instances at a bare minimum your application requires at all times.

Desired capacity

The desired capacity is the number of EC2 instances that Amazon EC2 Auto Scaling creates at the time the group is created. This number can only be within or equal to the minimum or maximum.

If that number decreases, Amazon EC2 Auto Scaling removes the oldest instance by default. If that number increases, Amazon EC2 Auto Scaling creates new instances using the launch template.

Maximum capacity

This is the maximum number of instances running in your Auto Scaling group, even if the threshold for adding new instances is reached.

When traffic keeps growing, Amazon EC2 Auto Scaling keeps adding EC2 instances. This means the cost for your application will also keep growing. That’s why you must set a maximum amount to ensure it doesn’t go above your budget.

3. Scaling policies

By default, an Auto Scaling group will be kept to its initial desired capacity. While it’s possible to manually change the desired capacity, you can also use scaling policies.

In the Monitoring lesson, you learned about CloudWatch metrics and alarms. You use metrics to keep information about different attributes of your EC2 instance, such as the CPU percentage. You use alarms to specify an action when a threshold is reached. Metrics and alarms are what scaling policies use to know when to act. For example, you can set up an alarm that states when the CPU utilization is above 70 percent across the entire fleet of EC2 instances. It will then invoke a scaling policy to add an EC2 instance.

Three types of scaling policies are available: simple, step, and target tracking scaling. To learn about a category, choose the appropriate tab.

Simple Scaling Policy

With a simple scaling policy, you can do exactly what’s described in this module. You use a CloudWatch alarm and specify what to do when it is invoked. This can include adding or removing a number of EC2 instances or specifying a number of instances to set the desired capacity to. You can specify a percentage of the group instead of using a number of EC2 instances, which makes the group grow or shrink more quickly.

After the scaling policy is invoked, it enters a cooldown period before taking any other action. This is important because it takes time for the EC2 instances to start, and the CloudWatch alarm might still be invoked while the EC2 instance is booting. For example, you might decide to add an EC2 instance if the CPU utilization across all instances is above 65 percent. You don’t want to add more instances until that new EC2 instance is accepting traffic. However, what if the CPU utilization is now above 85 percent across the Auto Scaling group?

Adding one instance might not be the right move. Instead, you might want to add another step in your scaling policy. Unfortunately, a simple scaling policy can’t help with that. This is where a step scaling policy helps.

Step Scaling Policy

Step scaling policies respond to additional alarms even when a scaling activity or health check replacement is in progress. Similar to the previous example, you might decide to add two more instances when CPU utilization is at 85 percent and four more instances when it’s at 95 percent.

Deciding when to add and remove instances based on CloudWatch alarms might seem like a difficult task. This is why the third type of scaling policy exists—target tracking.

Target Tracking Scaling Policy

If your application scales based on average CPU utilization, average network utilization (in or out), or request count, then this scaling policy type is the one to use. All you need to provide is the target value to track, and it automatically creates the required CloudWatch alarms.

. . . . . . . . . . . . . . . . . . . . .
Terabox Video Player