The Channel Angle is a monthly CRN guest column written by a rotating group of solution provider executives that focuses on the triumphs and challenges that solution providers face. If you are a solution provider executive interested in contributing, please contact managing editor David Harris.]
The technology sector has grown to be software-centric, and this new world requires a new way of thinking as more workloads are moved to the cloud. A cloud footprint can be configured through a provider’s console, but the problem is that anything manual has the potential for human error.
There are many compelling reasons why organizations should consider cloud automation. In the new world, companies can set up a cloud footprint using automation to reduce human error, make it quick and easy to create new environments, document setup for employee rotation, and do more with fewer engineers.
The process begins by creating the automation to start the servers and all the different components needed to run an application. Infrastructure automation, or software script, uses a configuration to create an environment. Once automation is perfected, it doesn’t change from one environment to another. It is one and done.
Once automation is written, setting up new environments allows IT to simply push a button. This avoids potential human error. It also makes it easy to quickly create additional environments based on business needs.
A performance test environment is an example of an additional environment that can be configured. Some application projects are large enough to require multiple user test environments, and automation makes it easy to create any additional environments.
Weaknesses with automating cloud deployments
Automation is a paradigm shift for both management and administrators, who are used to doing things manually. It requires skills that organizations probably don’t have and need to acquire. Also, the on-premises organizational structure doesn’t work in the cloud for most organizations, especially when doing automation. Conversely, people with automation skills often lack specialist knowledge in networking, security, DNS, and active directory, among other areas.
That said, the momentum continues to grow for organizations to have cloud footprints. For example, the CTO of a client of a large financial services firm has stated categorically that they do not want to make any additional investments in data centers or physical hardware. All new application workloads go to the cloud because the CTO wants to get out of the physical data center business and avoid the need to own property. Cloud automation is the only way for her to achieve her goal; manual administration would not scale properly.
However, some organizations feel that moving to the cloud will bring them cost benefits, which is illusory. Rather, the benefit lies in the speed to market and in allowing a company to become more agile and competitive. Automation is a prerequisite for achieving that speed. Otherwise, if you are simply moving your current applications to the cloud, you will move the problem to someone else’s data center.
When companies enter the cloud, they all make the same mistake of assuming that this is a technological change and that the business will be conducted in the same way, but with the new cloud technology. The problem with this mindset is that the cloud is all software-based and when workloads are automated, the old departmental or team silos don’t work. A silo could be a network administrator who is used to managing LANs and has no innate cloud skills.
On-premises technology skills don’t automatically translate to the cloud without a skill upgrade. So what happens is that DevOps teams that have cloud and automation skills end up having to add networking and other additional skills. This can leave local administrators feeling that their jobs are under threat and that they are excluded from the movement unless they are willing to acquire new skills.
Cloud automation isn’t about eliminating IT roles, it’s about changing them. Some humans have yet to figure out which rules make sense for their company and then someone has to automate them and push them out in order for them to be effective. In the cloud, security should also be automated.
The value of policy-based management
Ultimately, organizations are migrating to policy-based management, which establishes automated barriers that will prevent someone from doing things they shouldn’t.
This reduces the need to rely on people having to manually control systems. The main enemy of moving to the cloud is beginner mistakes. There have been instances where app developers have unintentionally opened databases so that they can be accessed over the Internet, simply because they don’t know any better. It’s a security breach if they do this because it increases the chance of your data being breached and seen by people who shouldn’t be seeing it.
A policy-based management plan for automating cloud applications and workloads will prevent rookie mistakes. Most cloud providers use APIs, so automating them is relatively straightforward.
The complexity of cloud automation will depend on the cloud maturity level of the organization. For customers just starting their migration, be sure to automate everything right from the start.
Organizations should make sure they set up procedures to automate their cloud footprint so they don’t have to implement anything manually. It gets a little tricky when a customer who has started with manual processes wants to adopt automation. So more work is needed to make it happen.
There are tools for doing this, but the process takes longer. But organizations that have automated manual processes are very satisfied in the long run.
A provider can teach organizations with a manual presence how to do automation as code. This saves a lot of time, effort and headaches.
Organizations embracing the cloud and automation will see savings in staff costs and dynamic scalability, and the ability to grow and shrink the cloud’s footprint based on demand. They will also see significant productivity gains.
Hybrid cloud presence even for large enterprises can be managed with relatively small staff when cloud automation is adopted. Compare that to the old world with an equivalent footprint, where everything has to be done manually and you’d be talking about a lot more people.
Mark McCoy is a managing partner and corporate hybrid cloud architect at
based in Chicago.