Cloud computing is boldly going where no other system has gone before. With so many organizations moving to some type of cloud platform, providers are finding more ways to create the true fully automated cloud environment. Now, we haven’t quite reached that point, but we’re getting closer.
Within the concept of cloud automation and orchestration are several important layers all working together. Starting at the data center level, the automation process includes technologies which, when combined, can produce some pretty powerful cloud infrastructures.
So, what are those cloud automation layers and why are they important?
- What you need to know. True cloud orchestration is being driven by open-source technologies. Unfortunately, there isn’t a lot of standardization here yet. But the technology and the development around it is developing very fast. For example, open-source IT automation tools such as Puppet from Puppet Labs or Chef from Opscode are now being used to automate management functions that previously required a lot of manual intervention. This heavy lifting used to be part of the administrative task. Now it can be completely off-loaded.
- Why it’s important. Yes, it’s a newer concept. But this certainly hasn’t stopped very large organizations from diving into the development and testing pool. Many are even using platforms like CloudStack, OpenStack, and even OpenNebula. BMW and Time Warner Cable, are two of the latest converts to OpenStack, for example. Throw big data into the mix and tie in data management solutions like MapR or Cloudera and you’ve got a very interesting field that's developing. Remember, your data is critical. And, creating powerful cloud automation services to replicate, secure and quantify this information is critical for you to stay competitive.
- What you need to know. Everyone is doing it. You should too. Many environments which work with cloud computing pretty much have to adopt a high-density computing model. Blade systems have come really, really far. So much so that you can basically insert a blade into, for example, a UCS chassis and let the hardware profile services do the rest. You are literally creating a puzzle-piece data center capable of amazing agility.
- Why it’s important. In a world of maximum efficiency, data center administrators simply don’t have the time to configure and deploy blades on an individual basis. With pre-built templates, administrators control entire data center blade deployments from one central console. Essentially, you’re reducing IT management while still improving workload delivery.
- What you need to know. This has been the savior for many cloud systems. Providers which host a cloud environment fully understand the dynamic nature of the platform. Anything can change in a second. Whether a server needs more resources or an application is getting pegged, circumstances can change in a heartbeat.
- Why it’s important. Automating ways to deliver applications and virtual servers can be a real life saver. Provisioning services are capable of spinning up VMs within seconds and are now also starting to allow users to connect to new resources. In working with intelligent load-balancing technologies, you’re able to intelligently provision and de-provision entire workloads on demand. Take VMware for example, which was faced with a dramatic increase in the volume of data, storage requirements, and the need for better management. To combat this, and as part of their software-defined data center, VMware turned to Puppet Labs (and gave them additional $30 million) to help manage IT resource much more efficiently.
The cost of moving to a cloud platform is shrinking, which means more organizations are capable of adopting some type of cloud model. Whether this is a migration to Office 365 or a full data center-based cloud deployment, the business drivers to move to the cloud are growing, and cloud providers are looking everywhere to increase efficiency. One great way to do this is a structured and layered cloud automation platform. Look for this technology to continue to develop – from the data center all the way to the cloud.