DevOps blog for beginner's

How to Implement Autoscaling in AWS: A Step-by-Step Guide

Autoscaling Your AWS Infrastructure for High Availability

Pawan Bhagat's photo
Pawan Bhagat
·

7 min read

In the previous blog, we saw the load balancer service, its type, and its practical implementation on the AWS cloud. The link is below.

https://hashnode.com/post/clf2jqoix000g09mh7sbwbc4l

So, some common challenges can be faced with load balancers in AWS: One of them is scaling.

Scaling: Load balancers can be a bottleneck when scaling up or down the number of instances behind them. Proper planning and configuration are required to ensure that the load balancer can handle the expected traffic volume.

Why Auto-Scaling?

In Load Balancer in AWS, its types, and practical implementation blog, we saw three web servers responsible for hosting a single website, which is managed by a single load balancer.

Suppose, if our website is a type of e-commerce website like Flipkart or Snapdeal and on some festivals or on-sale seasons there are lots of visitors who will visit our website for shopping.

If we have fixed numbers of Web-servers or a single web server then that server is not able to handle such type of heavy traffic, then there will be chances of a crash of the website due to heavy traffic and our website will be down.

So, to avoid such type situation we need extra web servers to handle such type of traffic and once the traffic goes down, we have to remove that extra web server from our infrastructure to better utilization of our resources and to save extra charges that we have to pay for web-servers

There are two ways to handle such type of situations

  1. Without the Auto-scaling: - Here we have to add and configure the web servers manually, this is not a recommended way because with manual intervention there will be a chance of getting errors.

  2. With Auto-scaling: - Here the web servers are added, configured, and removed automatically based on defined parameters.

Here, we understand the limitation of the Load-balancer, so we understood that along with the load balancer, we need auto-scaling.

This also provides the feature of auto-healing, which means if the issue is not with load or, unfortunately, if the instance failed for any other reason, it will automatically remove that failed instance and add a new instance with the same configuration within a fraction of a second.

How do we apply Auto-Scaling?

We have set the condition on a CPU threshold value

There are different mechanisms but one of the most usually is through the CPU utilization

A new instance will be created if the CPU threshold value is more than 80% of any existing servers.

And, if the CPU threshold value is less than 30% then the newly created instance will be removed.

We have to add such type of condition while the configuration of the Autoscaling group

Benefits and Features

  1. Fault Tolerance: - Detect when the instance is unhealthy, terminate it, and launch an instance to replace the old one.

  2. Availability: - Ensure that the application always has the right amount of capacity to handle the current traffic demand.

  3. Cost Management: - Save money by dynamically launching instances when they are needed and terminating them when they aren't.

Pre-requisite to creating the Auto-scaling

  1. Create the Load-Balancer: - This is used to manage the incoming traffic which is coming from the web.

    With the load balancer, we have to add an auto-scaling service.

    Kindly go through the below link to know more about load-balancer

    https://hashnode.com/post/clf2jqoix000g09mh7sbwbc4l

  2. Define the Launch Template: - It contains the details about the launch instance which is like below

    Which Amazon Machine Image?

    Which Instance Type?

    Which Volume?

    Security Group?

    Key Pairs?

    Which package needs to install? | Which are the web files?

  3. Define the Auto-scaling

Once the Launch Template is defined, we have to configure when the auto scaling will apply for that we have to know the below parameters

To which load balancer this newly created instance will be attached?

How many instances need to be launched (Minimum & Maximum)?

How to scale up and scale down the instances?

Which name needs to be given for instance?

Need to send the notifications when launching and removing instances

Practical Implementation of Auto-scaling

Step 1: - Create the load balancer

  1. Go to EC2 dashboard

  2. Click on Load balancer

  3. Click on create Load balancer

  1. Choose a classic load balancer

  1. Give a name to load balancer

  2. Create a new security group if you have already a security group then add the existing one.

  3. Configure Health Check

  4. Add EC2 instance

    Here don't add any instances, till now we have not created any instances just ignore it and click on next.

    (We will create the instance using AutoScaling)

  5. Add tags

    Apply tags to your resources to help organize and identify them.

  6. Review and create the Load balancer

    Here if you have to change or modify something on the load balancer you do it from here.

For more information about the load balancer and its implementation kindly go through the below link

https://hashnode.com/post/clf2jqoix000g09mh7sbwbc4l

Step 2: - Define a Launch Template

Here we have to pre-define the web server configuration, So whenever the web server is created using autoscaling it will launch using a pre-defined template

  1. Go to the Launch Template

  2. Click on create launch template

    Give Launch Template Name

  3. Choose AMI

    Click on Quick start

    Select Amazon Linux

  4. Choose instance type and Key pair

    Choose instance type which is free tier eligible.

    Create new key pair or use existing key pair (here I used my existing key pair).

  5. Security Group

    Create a new Security group or use the existing Security Group (Here I choose the existing SG which I created during Load balancer creation)

  6. Advance Details

    Write here the post-installation script for the Apache server in the user data section.

Once everything is done from the above steps, then click on Create Launch Template

Now, the Template is created successfully.

Step 3: - Defining Auto-Scaling Group

  1. Under Auto Scaling click on Auto scaling group

  2. Create on the Auto scaling group

  3. Give Auto Scaling group name and select template which we are created

  4. Choose instance launch options

For most applications, you can use multiple Availability Zones and let EC2 Auto Scaling balance your instances across the zones. The default VPC and default subnets are suitable for getting started quickly.

Here Add a subnet, So instance will be created on different subnets

  1. Configure advanced options

Choose a load balancer to distribute incoming traffic for your application across instances to make it more reliable and easily scalable. You can also set options that give you more control over health check replacements and monitoring.

click on Attach to an existing load balancer

choose from the classic load balancer

Select the Load balancer

  1. Configure group size and scaling policies

Specify the size of the Auto Scaling group by changing the desired capacity. You can also specify minimum and maximum capacity limits. Your desired capacity must be within the limit range.

Desired capacity: - (Always Greater than the minimum capacity) It means there are two servers on standby

Minimum capacity: - Minimum one server is running always

Maximum capacity: - (Always greater than the minimum and desired capacity), here maximum of 3 servers will create using Auto-Scaling

  1. Scaling policies

Target Value = 80

This means when the CPU utilization is equal to or greater than 80% then it creates a new instance using Auto-scaling

Instance need = 300 sec

This is for the duration between the creation of instance once after another

Then click on next, next and verify the details and click on create autoscaling group

Auto Scaling group is created successfully

The Total configuration for Auto scaling is done here,

Verification

  1. Go to EC2 dashboard and click on instance

Here, Automatically EC2 instance is created using pre-defined template with the help of Auto-scaling group

Note: - Both servers are created on different availability zones

  1. Click on EC2 dashboard and click on load balancer

    Copy the DNS address and paste on browsers

    on Browser

  2. If we deleted one instance then also we can access our website above link, Auto scaling not create instance because we configured minimum value is = 1 in auto-scaling group

  3. If we terminate all the running instance then autoscaling automatically create the instance as per Group size configuration minimum number of instance =1

  4. If the load exceeds 80% CPU utilization, it will create one instance, and if utilization remains high, it will create another instance until the CPU utilization falls below 80% (our configuration allows for a maximum of three instances to be created).

This is how we can use the Auto-scaling concept in our infrastructure also we see why not only a load balancer is sufficient to manage the load.

Note:- Once practice is done don't forget to delete the template, instances, Autoscaling group, key, and security groups.

Thank you for reading the blog!

DevopsAWSLinuxautoscalingCloud