Break down the IT silos that slow delivery of services

[JudyGoldman]

 Guest post by Eric J. Bruno

Enterprises need to break down the IT silos that slow delivery of services, but because most enterprise applications are rigid and too monolithic to support the agility and mobility requirements of the business, silos are the norm and services are slowed. By orchestrating the data center, enterprises can shift from managing infrastructure to managing services.

Using Agile to meet Agile's needs

IT is taking the very thing that's pushing them to move faster—the Agile movement—and using it to remove their own bottlenecks. According to Agile Methodology, Agile is an incremental, iterative process of software development consisting of small work intervals known as "sprints." An alternative to waterfall development, which relied on sequential development, Agile helps enterprises get to market faster because it focuses on continuous improvement. Rather than go to market with a complete, albeit frequently flawed product, companies can get to market faster with a minimum viable product that can be continuously improved.

Along with Agile development and, more recently, DevOps, automation has grown into a new buzzword, the long-awaited answer to the pain points of IT and a natural companion of Agile. Automation increases speed and accuracy by removing the manual work from IT processes, including updating and upgrading systems, responding to incidents and service requests, and provisioning operating systems. For most organizations, it results in an increase in the conversion of manual, error-prone procedures into automated scripts. Overall, this is a good development with many positive side effects.

From automation to orchestration

The key is to separate the main ingredients for a successful orchestration recipe: First, automate tasks, such as software provisioning and patch management; second, orchestrate processes, such as recovery scripts and server fulfillment requests. There are many factors in IT that have grown to require formal orchestration and control of resources, including:

• The use of virtualization across the board for compute resources
• The rise of software-defined networking and the software-defined data center
• The rise of the cloud and its ability to grow and meet demand as needed (otherwise known as elasticity)

With this, actual hardware, such as processing power (compute), networking and storage are treated as pooled resources, delivered to customers, and applied to demand as needed. This is how cloud elasticity is realized. These pooled resources are managed through software policies that include provisioning, security, backup and recovery, and other specific needs. In order to meet real demand on the fly, these all need to be in front and performed in real-time.

Orchestrating IT services, such as resource delivery, can result in faster delivery of critical services to the rest of your company. But it also means doing so consistently (with no surprises) while maintaining security standards, meeting compliance and regulatory requirements, and accurately predicting future needs. For example, with orchestration, IT is better enabled to solve problems as they occur (sometimes even predicting them before they occur) and adopt new technologies proactively instead of resisting change.

With orchestration comes self-service

Once the proper automation is in place, along with the needed policies around governance and security, you're in position to break down IT silos and enable self-service IT within your company. This includes the ability to provision new servers, deploy software services, and provide access to them across sets of users. This is essential when a hybrid approach to cloud and data center compute has become critical within your organization.

The key to successful self-provisioning includes the following considerations:

• Make sure it works across the full stack of IT services. This includes network infrastructure, compute servers, middleware and database software, storage systems, and even the custom applications your enterprise depends upon.
• Enable it across all environment. This includes traditional data center resources, the public cloud, private cloud instances, and hybrid/managed cloud implementations.
• Support the entire software lifecycle. This includes development, integration, test, release, and production operations.
• Provide access across the service delivery chain. This includes handling requests, provisioning resources for each request, and delivering needed services and support. But don't forget to include de-provisioning—automating the return of resources back to the "pool" for reuse.

One man's experience is another man's education

I was involved in an early instance of orchestration before it went mainstream. The company I worked for needed to build a web interface on top of a legacy application. The positive side of doing this was increased demand for their services, hence market and revenue growth. The challenge was enabling it without asking the legacy application environment to do something it wasn't built to do: grow elastically to increased user demand. The solution was to build a layer of adapter services around the older infrastructure to both shield and migrate services over time. Doing so in a compartmentalized, automated way enabled these building-block services to be orchestrated into entirely new applications.

The initial benefits of orchestrated service delivery include reduction in IT costs, increased efficiency, faster service delivery, and improved customer satisfaction—both inside and outside the company. But it also can be key in improving security and reliability, enabling new business models, acquiring new revenue streams, and reacting more quickly to changing customer and market demands. All without requiring the additional costs or new resources. When going into a project such as automating and orchestrating, it is a welcomed change to be able to not only see the expected results come to fruition but more so on the tangible, unexpected results that end up moving the needle.

To learn more about unleashing your IT potential, read the executive summary.

[darrenj1]

Im really trying to get OO in use at our organisation, but hitting stumbling blocks with teams not responding proactively to it, they just cannot see how a "ping" function for example is any better than when they do a ping via command prompt like now. They see the new OO processes as longer than current processes.

[JudyGoldman]

Hi, darrenj1,

Thanks so much for your comment.  Here's a reply from Dave Langlais at HPE:

Hi, darrenj1,, Automations and orchestrations have a pretty unique property in that the more you do, and the more complex, the value returned geometrically increases.  However, even very simple automation/orchestration - like the ‘ping’ function you mention - really can deliver value if leveraged.  For instance, it is true that doing a simple ping through the command line is really not very different than having a ‘ping’ function execute.

The real value comes when your teams realize that ‘they’ don’t have to do it. They can do other, more important actions. The ping function can do it based on conditional situations, logging results, analyzing trends/behaviors. And the results can be connected to other actions/functions to do ‘what’s next’ after the results of their ping comes back.  The power of automation/orchestration solutions like OO is when you start connecting individual actions together – that can be executed, error-free, anytime necessary, and always available.

It’s a huge leap in efficiency… We have many great examples of companies that have gotten past the point where you are.  We would be happy to connect you to others who have overcome the challenges you’re facing now. Feel free to contact me if you’d like more help.

Dave Langlais, HPE

dave.langlais@hpe.com