Create an Aurora Read Replica for an RDS Instance (Step-by-Step)

There are times when we want to scale up the existing databases deployed on the Amazon Aurora which is in house relational database written and created by Amazon itself from scratch. So today we are going to look at creating an “Aurora Read Replica of an RDS instance“.

Before going further, make sure you follow this -> MYSQL Database on Amazon RDS post because upcoming steps can only be performed when we have a running RDS instance, which we learned in that post, this post is in the continuation of that post.

Step 1: Create Aurora Read Replica

We have to select the RDS database instance of which we want to create aurora read replica. As can be seen from the image below, we first select the “geekylanedb” then from the “Actions” menu we have selected the “Create Aurora read replica” option.

Under actions select create aurora read replica — Under actions, select create an aurora read replica

Instance specifications

It is time to provide specifications for our “Aurora Read Replica” instance. We have provided the following details:

DB engine version = Aurora (MySQL)-5.7.12
DB instance class = db.t2.small – 1 vCPU, 2 GiB RAM
Multi-AZ deployment = Create Replica in Different Zone

Create Aurora read replica - Instance Specifications — Create Aurora read replica – Instance Specifications

Settings

A few more need to be provided, are as follows:

DB instance identifier = geekylane-auroraDB

Create Aurora read replica - Settings — Create Aurora read replica – Settings

Network & Security

We also have to provide some network and security related details as well:

Virtual Private Cloud (VPC) = Default
Subnet group = default
Public accessibility = Yes
Availability Zone = No preference
VPC security groups = Create a new VPC security group

Create Aurora read replica Network and Security

Database Options

Some database related details are required for creating “Aurora Read Replica of an RDS Instance“, those details are as follows:

Database port = 3306
DB parameter group = default.aurora-mysql5.7

Create Aurora read replica - Database Options — Create Aurora read replica – Database Options

Encryption

For this, we have to specify whether we want to encrypt the data on the “Aurora Read Replica DB instance” or not. As of now, we have selected the “Disable encryption” option.

Failover

Priority, here we have selected the “No preference“.

Backup

In this option, we are enabling the “Automatic Backups“, also we have specified about “up to how many days the Amazon should keep the backup“, we want it for only “1 day“.

Monitoring

We have enabled the monitoring for this instance, with the following options:

Monitoring Role = Default
Granularity = 60 seconds

Create Aurora read replica - Monitoring — Create Aurora read replica – Monitoring

Log exports

We left it all blank because we do not want any kind of logs to be logged.

Maintenance

Also, we do want the “Auto minor version upgrade” so we selected the “Yes” option. Moreover, we also have to specify the “Maintenance Window“, we have opted for the “No preference” option.

After doing all the things mentioned above, you will be seeing a button like shown below, simply click on Create read replica button.

Create Read Replica

Status = Creating

After doing all the steps, it is going to take a few minutes in order to create “2 instances“.

Note: Both the instances do have their own unique names, but they come under a unique name “geekylane-auroradb-cluster“.

Also, both the instances are going to be created in different, which is Amazon AWS intelligence.

Status = Backing-up

Before it creates the final cluster, the Amazon AWS is going to “back up” the original “Amazon RDS DB instance“.

From the picture below, we can confirm that.

Status = Preparing-data-migration

After backing up the original RDS DB instance, the instance is “Available” again.

Also, now the cluster is migrating the data from the original instance to the “Aurora Read Replica cluster“.

Original is available again but aurora is migrating — Original is available again but the aurora is migrating

Status = Available

From the screenshot below, we can see that the status of all the instances (original + new) has been changed to “Available“, which means that our “Aurora Read Replica DB instance” has been successfully deployed.

Step 2: Promote the Aurora Read Replica DB cluster

Now, we want to promote our newly created “Aurora Read Replica DB cluster“, and in order to do that, we have to select the “Actions” menu from there we have selected the “Promote” option, as shown in the picture below.

Promote database to be a read alone database

Step 3: Delete the original RDS instance

It is time to delete our original base RDS instance and to do that, again select the instance and go to the “Actions” menu and from there select the “Delete” option as shown in the screenshot below.

It is going to ask you for the confirmation on instance deletion, and a screen similar below is going to be shown to you.

Simply type the “delete me” and select the “I acknowledge….” option. After that click on the Delete button.

Status = Deleting

Firstly, it is going to create the original RDS DB instance and the status is going to change to “Deleting“.

Step 4: Stop the Aurora Read Replica Cluster

It is going to be a long step, and we want to cover each and every aspect of this step that is why we have provided images for each process-step.

First of all, in order to start the “Aurora Read Replica Cluster” process, we have to select the “Actions” menu and from there we have to select the “Stop” option, as shown in the photo below.

It is going to ask for the confirmation and you will be seeing a screen as shown below. Click on the Stop database button.

Status = Stopping

The status of the cluster has been changed to “Stopping“.

Soon, you will be seeing something as shown below. The status of both the instances in the cluster will be going to be changed to “Stopping” as well.

Status - all aurora read replicas are stopping — Status – all aurora read replicas are stopping

Status = Stopped

Notice that the status of the “Reader” instance is going to be changed to “Stopped” first.

After that, the “Writer” instance is going to get changed to “Stopped“.

Soon, the overall status is going to be changed to “Stopped“.

Full cluster has been stopped — The full cluster has been stopped

Note: Although we have stopped the Aurora Read Replica cluster in the previous step, in order to delete it, we have to first start it again. Also to confirm that, you can see the picture below for the reference, as it is not showing any “Delete” option for the cluster.