Are you equipped to deal with disaster?
As data and applications become increasingly critical for businesses, the need for quick recovery from disasters has assumed enormous significance. Any breakdown in the critical systems is likely to disrupt business transactions causing not just financial loss but even more valuable – dent in credibility. Fortunately, with rapid innovations from cloud providers including AWS, DR has become a lot more reliable and a lot less expensive.
Choosing your deployment site for disaster recovery
If your current system is in on-premise infrastructure, AWS can be a compelling option for its DR. AWS has a large global presence by having regions in the Americas, EMEA and Asia Pacific so you can choose a DR in a region that is closest to the geographical site of your system’s deployment. You can replicate your system (either in entirety or in parts) on the AWS and allow for smooth switching should there be any failure or disaster. Depending on business criticality, you could factor in ‘Recovery Time Objective (RTO)’ and ‘Recovery Point Objective (RPO)’ to arrive at an appropriate solution.
Recovery Time Objective: Targeted duration within which DR site should be operational.
Recovery Point Objective: Maximum acceptable period in which data might be lost in the event of a disaster.
AWS features facilitate a spectrum of DR solutions that suit the most demanding business needs and at the same time, cost a lot less as compared to how much they would have a few years ago.
Below, we look at four popular options that vary in cost and recovery time
- Backup and restore
- Pilot light
- Warm standby
- Multi-site
Backup and Restore
In this option, all essential data is backed up to AWS S3 at frequent intervals. On the occurrence of a disaster, the system can be restored in AWS using the most recent backup. The frequency of backups depends on acceptable data loss after a disaster occurs (RPO), which varies from business to business. Time taken to restore the application (RTO) depends on various services and the components that make up the application. With recent advancements in automation, both RTO and RPO can be significantly reduced. This option costs the lowest, as the only cost incurred is on the storage of backups.
Pilot light
In this configuration, you set up and run only the most critical parts of the system in AWS. Should there be a need for recovery, you quickly bring up the rest of the parts and scale up the production environment around the critical core. Even though this will cost more than the ‘Backup and Restore’ approach, if set up is right, this can considerably reduce recovery time (RTO). The data replication method that you employ is determined by the recovery point that you need.
Warm standby
A warm standby solution extends the pilot light elements. It further decreases the recovery time by keeping all required services always running. Business-critical systems are fully replicated on AWS and are kept ready to serve requests.
Multi-site
A multi-site solution runs in AWS as well as on your existing on-premise infrastructure in an active-active configuration. The requests could be load-balanced between on-premise infrastructure and AWS. A disaster in an on-premise system has no visible effect on end-users, as all requests get routed to AWS seamlessly (the only requests that could be affected are those in-flight on an on-premise system, which on re-submission proceed to AWS system)
Being in the market for over a decade, AWS has made available numerous services and features suitable for a wide variety of applications and computing workloads. CloudAccel has been helping customers benefit from the maturity and scale of AWS, including DR design and setup. While you continue to use on-premise infrastructure, having your DR on AWS can not only result in savings but also achieve better reliability.
