ADVANCED DEMO ON DATABASE MIGRATION USING DATABASE MIGRATION SERVICE

Migrate an OnPremises web application with a self-managed mariaDB database server into EC2 webserver and RDS-MariaDB database on AWS.

The on-premises environment is simulated using EC2 likewise the database. Original work by Adrian Cantrill

• States of this Demo
  • STAGE 1: Provision the environment using CloudFormation via a One-click deployment.
  • STAGE 2: Establish Private Connectiviy between the environments via VPC Peering
  • STAGE 3: Create and Configure the AWS side of the infrastructure
  • STAGE 4: Migrate Database and Cutover
  • STAGE 5: Account cleanup

STAGE 1

• Login to the AWS Console as user with admin permission and make sure to be on us-east-1 region. Click to provision the base infrastructure.
• Take note of the values of the following parameters
  • DBName
  • DBPassword
  • DBRootPassword
  • DBUser
• Scroll to the bottom, check the capabilities box and click Create Stack
• Wait for the stack to be CREATE_COMPLETE state
• Navigate to the EC2 console click on Running Instaces, the two instances CatWEB and CatDB will be running.
• Select the CatWEB, copy and open in a new tab the Public IPv4 DNS.
• This will load the web application running on the on-premises environment.

STAGE 2A - Create a VPC peering connection between the On-premises and AWS Environments

In production, a Direct Connect or a VPN used here would be used but here you are going to simulate it using VPC Peering to configure the connection between the two environments.

• Navigate to the AWS VPC console and click on Your VPCs
• In addition to the default vpc, you will see the awsVPC and onpremVPC running in IPv4 CIDR ranges 10.16.0.0/16 and 192.168.10.0/24 respectively.
• Creating Peering Connections
  • On the left menu, scroll down and click on Peering connections then click on Create peering connection
  • enter A4L-ON-PREMISES-TO-AWS as name tag
  • choose onpremVPC as VPC(Requester) and awsVPC as VPC(Accepter)
  • scroll down and click Create Peering Connection
  • both the Requester and Accepter VPCs are on the same AWS account, as such click on Actions then Accept Request

STAGE 2B - Configure Routing in each VPC to send traffic to each other

1. Configure Routes on the On-premises side VPC route by editing the route table

   • click on Route tables on the left menu of the VPC console
   • select the onpremPublicRT, click on the Routes tab and then click on Edit routes
   • click on Add route, copy and paste the IPv4 CIDR of the awsVPC
   • as Target, click on the dropdown and select ***Peering Connection to automatically select the peering connection created earlier.
   • click on Save changes

2. Configure Route on the AWS Side VPC by editing the route table there are 2 route tables on the AWS Side, (Private and Public)

   • select the awspublicRT and repeat the above steps to add a route using the onpremVPC IPv4 CIDR
   • repeat the process for the awsPrivateRT using the onpremVPC IPv4 CIDR At the stage we have configured gateway object and route table in both VPCs so both environment can communicate

STAGE 3A- Create the RDS Instance

• Navigate to AWS RDS console
• Create subnet group
  • click on Subnet groups on the left menu and click on Create DB Subnet group this is where RDS will launch the instances
  • enter A4LDBSNGROUP as name and **description
  • choose awsVPC from the dropdown and select us-east-1a and us-east-1b under **availability zones.
  • on the VPC console - subnets, there is a public and private subnets created earlier named aws-privateA and aws-privateB with 10.16.32.0/20 and 10.16.96.0/20 as IPv4 CIDR respectively
  • under subnets select the two subnets above (remember that our database is not accessible to the public therefore the instances should be launched in private subnet.)
  • scroll down and click on Create
• Creating the Database
  • Click on Databases - Create database
  • select standard create MariaDB as Engine option and free tier Template
  • goto the *parameters of the CloudFormation - Stacks console. there you will copy the parameter values to populate the DB Setting.
  • enter a4lwordpress as DB instance identifier & Master username
  • enter the DBPassword parameter value which is cats-dogs-rabbits-chickens as Master password and confirm
  • under Connectivity, ensure the option Don’t connect to an EC2 compute resource is selected. click on Virtual private cloud (VPC) dropdown and select awsVPC
  • ensure the subnet group a4ldbsngroup we created is selected and choose No for public access
  • select Choose existing for VPC security group (firewall) and select DMS-awsSecurityGroupDB-*
  • don't forget to remove the defaul security group
  • scroll down and expand Additional Configuration
  • use a4lwordpress for Initial database name
  • scroll down and click Create database

STAGE 3B - Create the EC2 Instance

• Navigate to the EC2 console
  • Click on Instances - Launch instance
  • name should be awsCatWeb and select Amazon Linux and the free tier Amazon Linux 2 AMI
  • set architecture to 64-bit(x86)
  • choose the free tier eligible instance type
  • choose a keypair or create new keypair
  • click to edit **Network settings and select awsVPC as VPC
  • select aws-publicA for Subnet
  • select Select existing security group and select DMS-awsSecurityGroupWeb-*
  • scroll down and expand Advanced details
  • under IAM Instance profile, select DMS-awsInstanceProfile-*
  • Click Launch Instance and then go to view all instances
• Install wordpress requirements
  • select the awsCatWeb instance, click on connect
  • on the SSH client tab copy the ssh command ssh -i "your_keypair_name.pem" ec2-user@ec2-52-87-209-171.compute-1.amazonaws.com
  • open you terminal in my case, a linux terminal and run the command
  • alternatively, you can connect using the Session Manager tab - connect
  • when you are connected to your linux machine run sudo bash for root permission.
  • run yum -y update to update the machine
  • run yum -y install httpd mariadb to install apache web server and mysql tools.
  • run amazon-linux-extras install -y lamp-mariadb10.2-php7.2 php7.2 to install php (a prerequisite for wordpress) and make sure apache is running and set to automatically startup should the instance restart.

  systemctl enable httpd
  systemctl start httpd
  

With this, we have a running apache web server ready to connect to our wordpress database.

STAGE 3C - Migrate wordpress content over

• first we need to edit the SSH config on this machine to allow password authentication temporarily so we can copy the wordpress data across to the awsCatWeb machine from the on-premises CatWeb Machine
• run nano /etc/ssh/sshd_config to open the file with nano
• scroll dow using you device arrow key to the line PasswordAuthentication no and change to PasswordAuthentication yes
• press ctrl+o the enter to save and cltr+x to exit the editor
• set a password on the ec2-user system user by running passwd ec2-user
• enter the same DBPassword, in production this should be different
• run service sshd restart or systemctl restart sshd to impliment those changes
• Navigate to the EC2 console and select the awsCatWeb instance and note the *private IPv4 address.
• then select the onpremises CatWeb instance right-click connect connect to the instance
• run sudo bash
• change directory cd /var/www/. You can run ls to see the files in the machine.
• run a secure copy to copy the html folder scp -rp html ec2-user@private_IP_of_awsCatWeb:/home/ec2-user and answer yes
• copy and enter the DBPassword on the parameter tab of the cloudFormation stack
• this will copy the wordpress local files (html folder) from CatWeb (on-premises) to awsCatWeb (aws)
• back to the awsCatWeb server
• do cd /home/ec2-user(same dir you are when you sshd) this is where we copied the html folder from onpremises to
• do ls -la to view the file
• do cd html
• run cp * -R /var/www/html/ to copy all the content recursive to the webroot of awsCatWeb
• Fix Up Permissions and verify awsCatWeb works
  • run the following commands to enforce the correct permissions on the files you have just copied

  usermod -a -G apache ec2-user   
  chown -R ec2-user:apache /var/www
  chmod 2775 /var/www
  find /var/www -type d -exec chmod 2775 {} \;
  find /var/www -type f -exec chmod 0664 {} \;
  sudo systemctl restart httpd
  

• go back to the ec2 console, copy the public IPv4 DNS and open in a new tab.
• If it is working, it means we have successfully migrated the web application from on-premise to aws instance which is still loading and pointing to the on-premises database. In the next stage, we will migrate the database webpage

STAGE 4 - Migrate the database from the CatDB on-premises database to the RDS-MariaDB provissioned in stage 3A above using DMS

1. Create subnet group
   • Search and click DMS to navigate to AWS DMS console
   • Click on Subnet groups on the left menu and then Create subnet group
   • provide A4LDMSSNGROUP as Name and Description
   • select the awsVPC from the VPC dropdown
   • under Add subnet click the dropdown and select awsPrivateA and awsPrivateB
   • click on Create subnet group
2. Create DMS Replication Instance
   • click on Replication instances on the left menu and click on Create replication instance
   • enter A4LONPREMTOAWS as Name and Description
   • select dms.t3.micro as instance class
   • select Dev or test workload(single-AZ) under Multi-AZ
   • scroll down to Connectivity and security ensure IPv4 is checked
   • select awsVPC under Virtual private cloud(VPC) for IPv4
   • ensure the a4ldmssngroup created earlier is selected under Replication subnet group
   • deselect Public accessible
   • click to expand Advanced setting and select DMS-awsSecurityGroup-* under VPC security groups
   • click `Create replication instance
3. Creat the DMS Source and Target Endpoints
   • Create Source Endpoint
     • click on Endpoints on the left menu and click Create endpoint
     • we are creating the source database endpoint so ensure Source endpoint is checked and *Select RDS DB instance is NOT selected
     • under Endpoint identifier enter CatDBOnpremises
     • Source engine should be MariaDB
     • select Provide access information manually
     • the Server name is the private IPv4 address of the on-premises database name CatDB in our running instance on the EC2 console.
     • MariaDB port is 3306 so enter this as Port
     • *username is a4lwordpress and password is the same *DBPassword value from the output parameter of our cloudFormation stack.
     • finally click on Create endpoint
   • Create Target Endpoint
     • click on **Create endpoint
     • select Target endpoint
     • check Select RDS DB instance because the MariaDB database is RDS-managed
     • most of the values will be automatically populated
     • scroll down and select Provide access information manually
     • provide the same password only as other fields will be automatically populated
     • scroll down and click on create endpoint
   • Test Endpoints
     • go to the RDS console and confirm that the Replication instance is Available status else wait
     • select any of the endpoints created
     • click on Actions - Test connections - Run Test
     • if successful the status will change from testing to successful
     • repeat for the other endpoint
4. Migrate Data
   • Create Database Migration Tasks
     • click on Database migration tasks on the left menu - Create task
     • Task identifier = A4LONPREMTOAWSWORDPRESS
     • select the Replication instance created
     • select the Source and Target database endpoints from the dropdowns
     • choose the Migration type in this case we are only migrating existing data
     • scroll down to Table mappings and ensure wizard editing mode is selected
     • click on Add new selection rule
     • click on Schema-Enter a schema just refers to the database name and the database table
     • source name we used while creating the database is a4lwordpress
     • use % as wildcard under Source table name to select all tables in the database
     • scroll down and click on Create task This will go through various state from creating - starting - Running and Load complete indicating a successful migration
5. Cutover the web application Instance to now Read/Write to the MariaDB database running on AWS
   • Navigate to EC2 console and select the awsCatWeb
   • ssh into the instance or connect through the session manager
   • run sudo bash to elevate your permission
   • cd to the directory /var/www/html which is where the web contents are
   • if you run ls command, you will see the file wp-config.php
   • run nano wp-config.php to open the file
   • goto RDS console, click to open the running database and copy the endpoint
   • scroll down to MySQL hostname change the IP 192.168.10.214 to the replication instance endpoint a4lwordpress.cb17d2jsxlse.us-east-1.rds.amazonaws.com
   • save and exit
   • Run the script below, to update the wordpress database with the new instance DNS name

   #!/bin/bash
   source <(php -r 'require("/var/www/html/wp-config.php"); echo("DB_NAME=".DB_NAME."; DB_USER=".DB_USER."; DB_PASSWORD=".DB_PASSWORD."; DB_HOST=".DB_HOST); ')
   SQL_COMMAND="mysql -u $DB_USER -h $DB_HOST -p$DB_PASSWORD $DB_NAME -e"
   OLD_URL=$(mysql -u $DB_USER -h $DB_HOST -p$DB_PASSWORD $DB_NAME -e 'select option_value from wp_options where option_id = 1;' | grep http)
   HOST=$(curl http://169.254.169.254/latest/meta-data/public-hostname)
   $SQL_COMMAND "UPDATE wp_options SET option_value = replace(option_value, '$OLD_URL', 'http://$HOST') WHERE option_name = 'home' OR option_name = 'siteurl';"
   $SQL_COMMAND "UPDATE wp_posts SET guid = replace(guid, '$OLD_URL','http://$HOST');"
   $SQL_COMMAND "UPDATE wp_posts SET post_content = replace(post_content, '$OLD_URL', 'http://$HOST');"
   $SQL_COMMAND "UPDATE wp_postmeta SET meta_value = replace(meta_value,'$OLD_URL','http://$HOST');"
   

   • press enter after pasting it to ensure the entire code runs

Navigate to the EC2 console and stop both intances running on-premises CatDB and CatWeb

• select each and click Instance state - stop instance Test that our web application still runs
• select the awsCatWeb instance, copy its Public IPv4 DNS and open in a new tab. If this loads, it means our web application now points to the RDS instance.

STAGE 5 - Cleanup You Account

1. Terminate EC2 Instances
   • select the awsCatWeb which you provissioned manually, right click and Terminate instance
2. Delete Database
   • navigate to the RDS console
   • select Database on the left menu
   • select the a4lwordpress click on Actions - Delete
   • uncheck Create final snapshot? and Retain automated backups
   • acknowledge and then type delete me
   • click on Delete me
   • wait for the process to complete before continuing
   • select Subnet groups on the left menu
   • select the a4ldbsngroup we created, click Delete - Delete
3. Delete the DMS Service
   • navigate to the DMS console
   • select Database migration tasks from the left menu
   • select the a4lonpremtoawswordpress task created, click Actions - Delete - Delete
   • click on Replication instances , select a4lonpremtoaws, click Actions - Delete - Delete
   • click on Endpoints on the left menu, select one of the endpoints created, click Actions - Delete - Delete
   • repeat for the other endpoint
   • if these doesn't delete, simply wait for the replication instace to finish deleting
   • click on Subnet groups on the left menu, select the subnet a4ldmssngroup, click on Actions - Delete - Delete
   • Navigate to the VPC console, select Route tables from the left menu
   • select one of the route created say awsPublicRT, click on Routes tab - Edit Routes, click on Remove to remove the route you added
   • click on save changes
   • repeate for the other two routes
   • click on Peering connections on the left menu
   • select the peering connection you created, A4L-ON-PREMISES-TO-AWS, click Actions - Delete peering connection
   • type delete then click on Delete to delete
4. Delete CloudFormation Stack
   • Navigate to the CloudFormation console
   • select the DMS stack created using the one-click deployment
   • click on Delete - Delete Stack. This will clean the rest of the resourses provisioned using the one-click deployment.

End of Project