Denys Rudyi - Fotolia

Manage Learn to apply best practices and optimize your operations.

Master the CloudOps model with four key steps

To evolve IT operations for cloud, organizations must embrace automation and change the way they think about resource provisioning.

The emerging term CloudOps -- short for cloud operations -- refers to how an enterprise runs and manages a cloud-based system. And, as many enterprises move applications and data off premises, there is a clear need to develop operational procedures around the use of public clouds.

When it comes to operating within clouds, enterprises often start from scratch. With public cloud, IT experiences a huge paradigm shift, in which the operation and administration of apps is done with limited visibility into the underlying infrastructure. CloudOps is part of this shift.

The four pillars of CloudOps

CloudOps is a new way of approaching IT operations, and is based on these four characteristics:

  1. It entails the use of abstract management layers to operate the cloud machine instances, storage instances, security, network and governance through a single pane of glass. This means administrators use a tool -- such as Apigee, ServiceMesh or RightScale -- to manage instances inside of the cloud, as well as other necessary systems, such as those for security and usage-based accounting.
  2. It involves a different type of provisioning. Instead of buying hardware and software to support an expanding set of applications, organizations manage the provisioning of machine instances on the cloud. There are two types of provisioning with CloudOps. The first is self-provisioning, where administrators allow cloud users to allocate their own machines, and then track their usage. But a more valuable approach is auto-provisioning, which allows the actual applications to request more machines and automatically deprovision them when not needed. However, administrators must track this model to avoid an expensive bill from their cloud provider.
  3. Instead of buying hardware and software to support an expanding set of applications, organizations manage the provisioning of machine instances on the cloud.
    It calls for limits. Organizations should set policies that limit what users can do within the public cloud. For example, while users can provision their own machines, they can't provision 1,000 at a time. Applications should also be limited, in terms of the capacity they can claim. This ensures applications don't use unbudgeted cloud resources.
  4. It requires automation. Organizations should automate as many processes as they can, including provisioning, user management, security management and application program interface governance. Automation means operational patterns are naturally repeatable, and allows a cloud environment to pretty much run itself. Also, consider automating the self-healing capabilities of a cloud provider. This means putting automated procedures in place to correct minor issues, such as machine instances going away, or network or data center outages.

CloudOps is complex. And, even as we move more workloads to the public cloud and learn more about development processes, CloudOps will become even more complex. That said, CloudOps should be cheaper and easier than traditional approaches, and require fewer resources. Moreover, we'll eventually be able to automate the CloudOps process so that IT staff just monitors the environment -- and that future is almost here.

Next Steps

Create rule sets for private cloud automation

Overcome these three hybrid cloud management challenges

Explore the relationship between DevOps and cloud

Dig Deeper on Cloud automation and orchestration