alphaspirit - Fotolia
The public cloud has emerged as a popular hosting alternative for some enterprise applications. The cloud promises rapid provisioning and vast scalability -- perfect for bursting resource-hungry workloads without putting undue strain on local data centers. However, hybrid and public cloud deployments have routinely been reserved for low-priority, test and development applications, or general-purpose business apps where some amount of downtime is tolerable.
Mission-critical workloads have typically been left on-premises, where dedicated administrators and infrastructure can ensure continued availability without compromise. This philosophy has started to change today as more organizations entrust critical applications to public and hybrid cloud providers. But before migrating a vital business workload to the cloud, it's important to weigh the risks and address the challenges involved.
Ultimately, computing is taking on the character of a utility, and the idea of hosting a mission-critical application in a public or hybrid cloud is becoming a more acceptable notion. But there are still serious hurtles that can torpedo the initiative in some situations. Don't rush vital workloads to the public cloud – instead, follow through on these five crucial steps before any migration.
Read every word of your cloud SLA
Never make a commitment to a cloud provider until you have read every word of the cloud service-level agreement (SLA). Your chosen provider should spell out specifics about uptime -- a vital element for critical workloads. However, it's not all about performance. Companies must also ensure that compliance regulations are met, and remember that this burden doesn't fall entirely on your provider. Don't put pen to paper until you're happy with the contract and are comfortable trusting your mission-critical apps to this provider.
Ensure mission-critical app performance
You can't risk having too much downtime for your high-priority workloads. But turning over your applications to a cloud provider doesn't take the onus off your company for Performance monitoring. Always use tools and procedures to make sure you're happy with your app's performance.
Keep cloud costs in budget
It costs money to keep mission-critical apps running, particularly if you don't do your due diligence before migrating. During the cloud planning stages, look into your application's resource usage, the CPU, memory, disk and network I/O. Then, take the steps to make sure that these variations won't break your budget when aligned with a provider's pricing structure.
Prepare for disaster to strike
As IT pros have learned again and again, you must always expect the unexpected. Disaster can strike even for the strongest cloud environment. This is why, in order to keep your critical workloads running, you need to ask the right questions about your cloud provider's disaster recovery protocols.
Know how to break up with your cloud provider
Sometimes things just don't work out. There are many reasons to end a relationship with your cloud provider, but it's what happens next that's most crucial. Always give yourself an out, particularly when you're working with mission-critical apps, and be ready to return the application in-house or move it to a better provider as needs and opportunities dictate.
About the author:
Stephen J. Bigelow is the senior technology editor of the Data Center and Virtualization Media Group. He can be reached at [email protected].
Cloud application migration under the microscope
Performance monitoring for cloud superior to virtualization?
Security improvements highlight new cloud models, IoT