A move to a public cloud often manages to drive down capital expenditures. If it’s a public cloud where you’ll be running applications that you have to size for performance, there are some skills you’ll want to invest in before you make the move – principally:
- Monitoring
- Security and Compliance
- Orchestration & Automation
Monitoring the performance of your applications is critical. Standard monitoring practices you’ve been comfortable with won’t necessarily work well. For example, you’ll need to adjust your latency and response time metrics based on the characteristics of the datacenters you’re using. As operators of our own (Seagate) and other clouds, we’ve found that we’ve had to adjust our operational templates to allow for up to 3x longer latency and response time to set reasonable targets for some 3rd party clouds we manage. Be prepared to monitor the consumption of multiple cloud resources along with upload and download bandwidth utilization and tie these to operational costs. Often, you’ll find that you are being ‘nickel and dimed’ for specific services.
Security and Compliance expertise must be developed beyond what you’ve had in place for your own datacenter. You’re essentially exposing your entire environment to the outside world when you run your business in the cloud and need to ensure that you have properly isolated systems and data access. Unless it is built into your application, you’ll find you have less control over security. Be ready to deal with the reality that you won’t be able to make database calls in the cloud and, unless you invest in scripting, you’ll end-up with less access. Needless to say, you will want to narrow down the number of access pathways and reduce the number of ports open, potentially to a small IP range.
When you move to the cloud, you’ll often hear that you don’t need to invest in sizing it right up front. In fact, you want to be careful about this, just as you are in your own datacenter. If you’re putting an application in the cloud on your own, be sure to design an N Tier architecture along with orchestration to scale up and scale down automatically.
Orchestration and Automation is therefore critical to taking advantage of cost savings when you can scale down, and to reduce the human cost of setup and configuration on-the-fly as your business needs change. Often, this is not a skill-set you will have in-house so you’ll need to contract with an engineer that is skilled in working with the APIs and resources in the cloud you’ve selected or permanently hire that skill-set into your team.
The amount of preparation and skill-set building need prior to moving to the cloud is complex. It is directly proportional to the level to which you’re selecting a cloud/application combination that is pre-designed to work well together vs. a selecting a cloud infrastructure where you will have to do the performance tuning of all resources after you design your application deployment. Seagate’s Cloud Backup Service, for example, falls in the camp of an application tuned specifically for our clouds (Seagate, Azure, Managed Service Provider). Purchasing a software application and then deploying in a compute oriented public cloud would fall in the latter camp, requiring sizing, tuning and robust on-going monitoring to ensure performance and cost control.
For an enterprise-class organizations, making this investment makes sense. On the other hand, due to the orchestrations and optimizations involved, smaller businesses should expect that it won’t be as easy to eek-out every last dime of cost savings.
Originally published here.