This topic describes the different disaster concepts that are referred to in Druva disaster recovery.
Disaster Recovery concepts
This topic lists the Disaster Recovery-specific concepts.
Druva AWS proxy
Druva AWS proxy is an Elastic Compute Cloud (EC2) instance that runs the Druva disaster recovery service. It orchestrates copying data from the Druva Cloud to your AWS account and creates a DR copy at a frequency specified in the DR plan. The Druva AWS proxy runs in your AWS account. The Druva AWS proxy is launched in the same AWS region where the virtual machine backups are located. The EC2 instances are started in the same region for disaster recovery.
Disaster recovery plan
Through a DR plan you can preconfigure various aspects of disaster recovery for a single-click failover in the event of a disaster. DR plan encompasses virtual machines configured for disaster recovery, the AWS account used for disaster recovery, the replication frequency, the network settings, instance-related failover settings, and the order in which you want to recover the instances.
The Druva AWS proxy converts a backed up virtual machine recovery point into a writable snapshot called the DR copy. The Druva AWS proxy then stores the DR copy in the AWS account. At the time of a disaster when you run a failover job using Druva, the Druva AWS proxy:
- Creates an EBS volume using the DR copy
- Injects the drivers into the EBS volume that are required to boot the EC2 instance
- Starts the EC2 instance using this EBS volume
This EC2 instance serves as the replacement of the virtual machine which is experiencing downtime during a disaster. The DR copy is updated based on the replication frequency you specify at the time of creating a DR plan.
The instance type determines the hardware of the machine that you use for your instance. A wide array of instance types provide various combinations of CPU, memory, storage, and network capacity for different needs of the application that you plan to run the instance on. For more information, see Support matrix and Manage failover settings.
Recovery Point Objective
Recovery Point Objective (RPO) is the point in time in the past to which you can recover data when a disaster occurs. In other words, RPO defines the amount of data the organization can afford to lose during a disaster.
For example, the organization has an RPO of 24 hours, and the backup is scheduled daily at 8 PM. If a disaster occurs at 7.59 PM, you can recover data that was backed up on the previous day at 8 PM. However, you lose the data generated after the last backup.
Recovery Point Actual
Recovery Point Actual (RPA) is the amount of data an organization actually lost due to the failover that is run at the time of a disaster recovery.
For example, an organization has an RPO of 24 hours, the virtual machine backup is scheduled daily at 8pm, and the replication frequency in the DR plan is set to Immediately after backup. At 8pm, on August 24th, 2018, Druva takes a backup of the virtual machine, and updates the DR copy of the virtual machine in the organization's AWS account. At that point, the virtual machine held 8 GB of data. On August 25th, 2018, at 11:00am, the virtual machine holds 9 GB of data. If the virtual machine fails at 11:15am on August 25th, 2018, and the organization's administrator runs a DR failover, the virtual machine is failed over to the EC2 instance that is created using the August 24th, 2018, 8pm DR copy which serves as the recovery point. The EC2 instance that is created after failover holds 8 GB of data, and 1 GB of data is lost. The 1 GB of data that is lost after failover is the RPA.
Recovery Time Objective
Recovery Time Objective (RTO) is the amount of time you set for your application to recover from the point-in-time the disaster occurs. In other words, RTO defines the maximum tolerable outage. The application must be restored within its defined RTO to ensure business continuity.
For example, the organization has an RTO of 30 hours. When a disaster occurs, the organization must recover all its critical IT services within 30 hours from the point in time the disaster occurs.
Recovery Time Actual
Recovery Time Actual (RTA) is the amount of time Druva takes to failover a virtual machine to an EC2 instance running in the organization's AWS account after you run a failover job on the Management Console.
The frequency at which Druva updates the DR copy of a virtual machine based on the available virtual machine recovery point in the Druva Cloud. For example, if the replication frequency is defined as immediately after backup, the DR copy of the virtual machine is updated as soon as Druva creates its latest recovery point. For more information, see Replication Frequency.
This section lists the AWS-specific concepts that Disaster Recovery leverages.
Amazon Web Services Account
An Amazon Web Services (AWS) account allows you to establish a formal relationship with AWS. With the AWS account, you can access the AWS account resources and Web services to the fullest. When you sign up for AWS, the AWS account is automatically signed up for all services in AWS. You are charged only for the services that you use.
Amazon Machine Image
An Amazon Machine Image (AMI) is an encrypted machine image stored in the Amazon Storage, such as Amazon Simple Storage Service. For more information, see Amazon Machine Image.
AWS CloudFormation enables the AWS resources deployment process by automating the creation and management of the AWS resources in the AWS environment.
For more information, see AWS CloudFormation Concepts.
AWS S3 bucket
AWS S3 bucket is the cloud storage provided by Amazon to store customer’s backed up data. Customers create their AWS accounts in their AWS S3 buckets that act as a secondary site for Disaster Recovery.
For more information, see Working with Amazon S3 Buckets.
IAM Policy is a document that allows you to define permissions for users, groups, roles, and resources.
An IAM Role is an AWS identity that provides access capabilities to AWS users. The role is used to delegate access to users to use AWS resources. The IAM Role ensures that the Druva AWS proxy has sufficient privileges to import data from the Druva Cloud to the AWS account and create an AMI or DR copy based on the Druva AWS proxy version.
For more information, see Amazon security best practices in IAM.
A VPC is a section of the AWS Cloud specific to the AWS account. You can launch the AWS resources, such as EC2 instances, in the virtual network. For more information, see Amazon Virtual Private Cloud.
A subnet is a segment of the IP address range of a VPC to which you can attach the EC2 instances. Subnet groups the instances according to the security and operational needs.