Disaster Recovery and Azure ASR
Disaster recovery (DR) is a critical area for an organization that allows them to maintain or quickly resume mission-critical functions following a disaster. There are two general measurements each organization has to measure how fast they can recover and how much data they can afford to lose.
Of cause, you might be wondering whether isn’t it obvious to have systems in place to recovery automatically and not to lose any data at all right?
Yes, it makes sense to have to systems in place to achieve the above but it also a matter of costs involved to set up such a system which needs to align with business objectives.
This is where the following measurements are important based on organization strategy for DR.
- Recovery Point Objective (RPO) — Is considered as the maximum targeted period in which data might be lost following and disaster or due to a major incident.
- Recovery Time Objective (RTO) — Is considered as the maximum duration of time it takes to restore the systems after a disaster.
Disaster Recovery Approaches
Based on RTO and RPO, systems can be implemented to restore the business operations following a disaster using variations and combinations of the following reference approaches.
- Backup & Restore — As the name suggests, this involves taking backups of the running system ( e.g Databases, Files, Virtual Machine Images) and restore in a different physical location following a disaster. The frequency of the backups are considered based on the RPO and will take more time to restore impacting the RTO.
- Pilot Lite — In this scenario core components (e.g databases) of the environment is always running in another location with automation and configuration in place to quickly provision the rest of the environment.The idea behind pilot light comes from the analogy of gas heater where a flame that’s “always on” can quickly ignite the entire furnace to heat up a house.
- Warm Standby — This is a scaled-down version of a fully functional environment that is always running in a different physical location. This further reduces the recovery time.
- Multisite — This is a specialized configuration where both environments are used in an active-active configuration. The application traffic is always distributed across these environments. It allows to automatically handle DR situations with auto-scaling capabilities for the running environment to handle the excess load.
Implementing these scenarios can be challenging without having the necessary infrastructure, tools, skills, and investments. However, using Cloud infrastructure has proven to be more effective in addressing most of these concerns.
Using Cloud for Disaster Recovery
Generally, setting up a fully functioning disaster recovery solution requires a high commitment in term of costs for the physical data centers and hardware. With the advancement in Cloud infrastructure, the upfront commitment in setting up a disaster recovery solution has gone down significantly. This mainly due to the on-demand resource provisioning and billing models provided by Cloud vendors such as AWS and Azure.
In addition to the on-demand infrastructure, their many utilities provided by Cloud to automate the process of DR. Most of these utilities are service level configurations (e.g Database Replication, Storage Level Replication, Virtual Machine Snapshots). However, having a centralized management and migration utility for DR is important to implement,
- On-Premise to Cloud.
- Across Cross Cloud Providers.
- Same Cloud Provider multiple Regions.
One of the key services that centralize the management and configuration of DR is Azure ASR (Azure Site Recovery) which facilitates to implement any of the above scenarios.
Azure Site Recovery
Azure ASR falls into DR as a Service (DRaaS) solution that helps ensure business continuity by facilitating to operate the business systems and applications functional even parts of the system fails due to a disaster or outage.
Azure ASR provides configurations replicates workloads running on physical and virtual machines (VMs), In a Cloud provider like AWS or in Azure itself from a primary site to a secondary location (In Cloud from one region to another). When an outage occurs at your primary site, Azure ASR ensures that the secondary location is ready to take over the application traffic. It will gracefully fail back to the primary site once it’s up and running again.
By using Azure ASR, it greatly simplifies the configuration effort to implement a DR solution also providing the abilities to simulate DR situations, centralize DR management and align with business strategies to ensure that the DR solution functions as expected.
When to use Azure Site Recovery?
If you have an on-premise infrastructure or has virtual machines running in any cloud provider, Azure Site Recovery is a great tool to replicate them across multiple infrastructures for DR. It supports the major virtualization environments, storage infrastructure. For more details on features supported by Azure ASR, refer the site recovery support matrix to Azure.
If you are using a different cloud provider like AWS, Azure Site Recovery provides the bare minimum capabilities for DR (For IaaS counterparts). This is because there are various Managed Services that offers support for DR natively in AWS which is not supported by Azure ASR. It will be the same if you are using Azure Managed Services and planning to replicate across regions. Therefore when having Cloud Region to Region or Cross Cloud DR solutions where there are more Managed Services or Native Cloud Services involved, Azure ASR won’t be the right fit.
