The cloud has become an important asset for businesses of all sizes, and most organizations have implemented some type of hybrid cloud environment in which the cloud is used for several different IT functions, including development, testing, backup, and business continuity. For many companies, these cloud resources have become an integral part of their infrastructure and they have leveraged the hybrid cloud for the high availability and disaster recovery (HADR) of their on-premises virtualization infrastructure. There are several important advantages to implementing HADR in the cloud. The cloud is an architecture that provides extremely high uptime and its ubiquitous connectivity enables cloud provider resources to be accessed globally. Perhaps more importantly, the cloud also offers a pay-as-you-go model in which you pay only for the cloud resources that you use. This can make the cloud a much lower cost option than a separate physical HADR site.
Ensuring the availability of hybrid cloud resources has different challenges than implementing on-premises HADR implementations. A common mistake that businesses make when they modernize their heterogeneous IT environments to incorporate the cloud is that they often fail to make corresponding enhancements to the protection environment as well. When the corresponding protection modernization efforts are not implemented, virtual machine (VM) availability can be limited and recovery failures can result.
Some of the important factors that you should consider when you are modernizing your hybrid cloud availability environment include:
- Understanding Your SLAs – One of the first things that you need to do is understand the SLAs that you are designing your high availability environment for. The SLAs required for your organization drive the HADR solutions that you need to implement.
- Setting Your Replication Intervals – The replication frequency that is used between your on-premises infrastructure and the cloud has a big impact on the Recovery Point Objectives (RPOs) that you can achieve. More frequent replication intervals enable lower RPOs, but they also typically require more resources and more network bandwidth.
- Minimizing WAN Latency with Data Compression – One of the biggest factors that influence replication is the amount of data that you sending from your on-premises infrastructure to the cloud. Taking advantage of technologies such as WAN data compression can enable you to create more efficient VM replication and to possibly increase your replication frequency.
- Providing Multiple Data Recovery Points – Media corruption is one of the most common sources of recovery failure. When you design your HADR solution, make sure you have multiple data copies to restore from.
- Testing Your HADR Solution – It’s not enough to create an HADR solution and just hope it works. To ensure high availability, you need to periodically test that solution to confirm that it actually works as planned.
- Taking Advantage of Low Latency Direct Connect Solutions – Finally, if your business requires low latency, higher performance solutions beyond what the Internet can provide, then direct cloud connection technologies such as Azure ExpressRoute and AWS Direct Connect can enable you to bypass the public Internet and make direct connections to your cloud provider.
The cloud can be an effective HADR asset to your company, but to take full advantage of it, you need to make sure the availability technologies you implement keep pace with your hybrid cloud infrastructure.