UPDATED 01:10 EDT / FEBRUARY 23 2016

NEWS

8 tips you should know before re-architecting IT

Over half of enterprise IT projects fail, resulting in staggering amounts of wasted time and money. However, it doesn’t have to be this way. There are easy steps every enterprise can take to ensure success. Before you take on an enterprise online re-architecture project, here are eight lessons to keep in mind.

1. Find a business transformation catalyst

Think of enterprise architecture like a car. If you never replace the oil, the engine will burn out and you’ll find yourself stuck with an expensive overhaul. But if you pay for maintenance as you go, it will ultimately cost you less because you’ll keep the engine in sound operating condition.

The lesson: Don’t wait three years to do something you should have been doing all along. Your enterprise architecture should always be evolving. Don’t allow systems that need fixing to remain under sustaining engineering teams until it is too late.

A business transformation catalyst is an application design and development provider that drastically increases the value of engineering. In other words, it’s a partner that constantly keeps your enterprise software evolving to match both your business goals and the larger competitive landscape.

2. If you’re painted into a corner, don’t stop

Most large enterprises have a lot of custom applications and as a result, find themselves painted into a corner. Their portfolio has so many customized, interconnected components that the cost and complexity of adding functionality becomes enormous. This slows down everything. Failure to continuously evolve the portfolio only causes the problems to stack up over time.

It may seem like the architecture is so broken that the only solution is to tear it down and start from scratch. But if you wait to go live until that whole rewrite is complete, you’ll paint yourself into a corner again because you’ll never be able to future-proof yourself. By the time your rewrite is complete, your business’ needs will have changed again. There are ways to get out of the corner without tearing down the whole house.

3. Tear down vertical team silos

The problem with vertical team silos is that no one team has a complete view of what is going on. Each team is responsible for one specific function and this can make it difficult to deploy anything new because it requires changes across every one of those silos. This slows deployment down and comprises agility. Vertically configured teams also create dependency between teams, which means if one service goes down, every service is affected. Interdependent teams create brittle infrastructure, and when services are tightly coupled to one another, this can cause serious performance issues.

For example, every individual service may have 99% uptime, but when you factor that 1% across hundreds of services, the total uptime SLA guarantee could end up closer to two-thirds.

Today’s enterprise architectures require a different, more horizontal, service-based infrastructure. Decoupling services is a powerful way to streamline business processes and optimize performance. Loosely coupled, resilient units of functionality provide more stability and agility because enhancements to services can be pushed to production independently of each other. And because service-based infrastructure allows every team to work asynchronously, they aren’t affected by performance issues in other teams.

4. Centralize collaboration

Businesses today are increasingly global, with teams around the world operating across multiple time zones. This causes problems. For example, getting a question answered by a team 10 time zones away can delay a response by 24 hours. That creates miscommunication, delay and misunderstanding.

The most effective way to navigate multi-site/multi-country development is to organize teams around specific projects, rather than trying to grab resources out of an agency style pool where there may be multiple, simultaneous projects that slice across teams. Establish clearly defined delivery teams centered around the actual work stream. Enable them to see the primary axis of customer value delivery from end to end. If these teams are collocated, then the flow of communication is more timely and of significantly higher quality. If they’re geographically distributed, use real-time collaboration tools shared information repositories.

5. Get a fresh set of eyes

Focusing exclusively on a specific problem limits your sense of perspective, which makes it difficult to know if you’re making progress. Certain functional aspects may be met, but the overall project is still a failure.

A fresh set of eyes is essential for providing an objective, bigger-picture perspective, as well as a cold dose of reality when needed. Enterprise organizations undertaking an IT reconstruction should call in experts who haven’t been involved in the project and who can clearly articulate where the problems are. These outsiders can assess where the key pain points are, what is working and what isn’t. They can also identify inefficiencies, help set priorities, and gauge how well the project is moving forward.

6. Preserve the pieces

A fresh set of eyes can also be instrumental in helping an organization figure out what elements to keep. As noted earlier, a rewrite doesn’t mean require replacing the entire system. Rather, you should preserve the pieces that do work and put them together in a new, more effective way. Often, more pieces are harvestable than people realize. For example, tackle a new application tier, but leave the database tier alone. The old and new system can co-exist because they share a common data model, and then you can gradually bring functionality up on one side and down on the other.

Harvesting existing assets also reduce risk. Lots of little steps are better than one giant step, and will prevent you from wasting time, money, and effort. You’ll also learn how to efficiently reconfigure existing services as new business demands come into play, which requires far less time or money than coding something new.

7. Plan your rollout strategy

One of the major challenges of re-architecting is to ensure that performance is not compromised during the process. Decoupling services so they function asynchronously prevents downtime and increases responsiveness. One service going down won’t have a domino effect. It also allows organizations to prioritize those services that are more important than others and to deliver independent service updates.

Education is also an important part of rollout. Plan to train development teams about the rationale behind the changes, go over new technology and walk through a few of the endpoints so they can take over.

8. Validate, validate, validate

You should never wait 18 months to get value from anything, whether it’s dieting or re-architecting your systems. Validate at every step along the way.

Start by validating your architectural assumptions. Are the sound? This step will help you weed out small errors in judgement before they become big mistakes.
Secondly, validate your SLAs. Don’t wait too long to confirm that you will be able to meet your nonfunctional SLAs from end-to-end, not just at the service level. This ties into the third key thing to validate — the client/customer experience. Ultimately, what matters most is giving the client/customer what they want at every step. Don’t wait until a customer has problems to realize that they exist.

Every enterprise will have to update its online architecture at some point. The market is too competitive and things move too fast to tolerate outdated or underperforming systems. The good news is that following a few simple steps can dramatically improve the efficiency of a re-architecture project and avoid unwieldy and expensive failures.

photo credit: billfrog2 via photopin cc

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU