For more than 65 years, TransCanada has proudly delivered the energy that millions of North Americans rely on to power their lives and fuel industry. We are leaders in the safe and reliable operation of pipelines, storage facilities and power-generating stations across Canada, the U.S. and Mexico.
In the energy business, we are committed to supporting long term strategies but always need to be ready to react quickly to immediate business needs. For TransCanada, that means planning and building relationships in communities that last decades. At the same time, we need to think in the moment to react to short-term business conditions influenced by everything from weather to politics to usage patterns.
Our move from data centers to cloud provides an interesting look at this short- plus long-term approach. The project began as a data center consolidation plan – not leveraging the cloud for re-hosting. Over the years, we have accumulated significant data center capacity from company growth through acquisition. So the long-term plan was to reduce our data center footprint and cut costs.
In the process of studying the issue, we looked at ways to offload a small percentage of workloads to the cloud. The cloud benefits were compelling and soon we were modeling out more intensive cloud scenarios and ultimately decided to go almost all in. Today, we are shrinking our data center use down to a bare minimum for real-time systems only and moving all other workloads to the public cloud.
In other words, we made a quick-turn decision to change the course of a long-term plan. We are still cutting costs and promoting efficiencies, as we planned early on in our data center consolidation initiative. But now we are adding additional long-term benefits: better security, more scalability, increased reliability for all systems and a newfound ability to apply analytics to a wide range of business operations.
A Fast Start
Although our cloud transformation journey is still in process and will take a while to complete, it did start quickly. In January 2017, we began our data center consolidation analysis. By March, we had decided to move 30 percent of our workloads to the cloud. By May, we opted for a complete shift, moving all corporate enterprise workloads to the cloud, keeping a reduced data center footprint for mission-critical Supervisory Control and Data Acquisition (SCADA) operations. We chose a service provider in July, and by early 2018 we were starting to migrate apps.
The first decision we made was to go with IaaS rather than PaaS. There are advantages to doing a full-on rewrite of your applications using PaaS — you get all kinds of flexibility and can innovate to fully optimize each app. But this is a huge commitment right out of the gate, and we felt we would be better off with an IaaS approach, lifting and shifting each of our apps to the cloud over the course of a 2 years.
The lift-and-shift approach was easier to sell to business users. They would get the benefits right out of the box — all the agility, speed and reliability associated with cloud infrastructure platforms. Users would be able to leverage data links immediately without having to clean up the data. Plus, IaaS buys us time to train and transform our organization. The idea is to get off the bare metal as fast as possible, migrate the legacy apps to IaaS, then rewrite them in the future if required.
We considered a number of factors in our choice of AWS as our cloud provider. We looked at the cost, capabilities, maturity of the platform and toolset offered, and AWS came out on top in each category. As part of our process, we gave each potential vendor a set of applications and asked them to demonstrate how they would move the apps from our on-premises environment to their particular platform.
Help With Migration
Next, we needed a partner to help us move through the migration stage of our cloud journey. Based on recommendations from AWS, we engaged Cloud Technology Partners (CTP), a Hewlett-Packard Enterprise company. We were impressed with the methodology, tools and resources CTP brought to the project. CTP helped us classify our apps into multiple categories based on each app’s readiness to be hosted in a cloud environment, and then created a schedule to migrate the apps onto the AWS platform.
By mid-2018, we had migrated about 20 percent of our apps to the cloud environment. Overall, the migration has run smoothly. Migration approaches under an IaaS model have resulted to date in minimal disruption and outage time for business operations. Perhaps the biggest challenge has been gathering the documentation for hundreds of applications, to make sure we weren’t missing anything. Then there is the issue of getting all the right teams together to test and support an application once it has moved over. Keeping the pipeline full with apps by adhering to schedules has been critical to the performance of the project.
Training the staff has been one of our top priorities – and one of our greatest successes. We are essentially reinventing ourselves to work with the cloud. We wanted our core staff – core employees who are shareholders in the company – to embrace technology change, so we put together programs to bring them on board.
Offering a blend of traditional in-class and self-service training, we set a target of having 5 percent of the staff achieving basic AWS certification by the end of 2018. Midway through the year, we are at 7 percent. Employees are embracing the different programs – everything from day-long immersion sessions, to study groups, to hackathons. One of their favorites has been a game called “Zombie Apocalypse,” where users try to save humanity using AWS tools. This has been a spirited move for a relatively conservative corporate culture.
We are starting to see the cloud initiative generate some of the benefits we had forecast. Our plan calls for cost savings to be achieved over a five-year period. We are starting to reinvest financial capital into programs that will add business value. A security assessment we conducted leveraging third parties concluded that we are more secure in the cloud than we were on premises. Our time-to-market goal — being able to provide services to business faster – has been proven out through our innovation series and three business validation tests. With regard to “full reliability,” you can do that replication seamlessly in a cloud environment.
Commitment to Success
We have an exciting and aggressive journey ahead. Ensuring we have effectively implemented cost management throughout the process to support our business case, continuing to migrate critical applications and maturing posture through reliability, and transforming our team’s capability to operate and innovate through technology are all fundamental to support TransCanada’s business strategy.
For companies in our industry, cloud transformation journeys can be long and drawn out. Our journey is still relatively new, and we are still working to integrate all the capabilities. But we are developing good momentum. To generate successes early on, you need to have a solid strategy, a good program and excellent partners. You have to create urgency within your organization, because you need to have everybody on board. Using this set of tactics, you can position a cloud project for both short- and long-term success.