Contributing writer

Middle Eastern CIOs share lessons learnt from cloud migration projects

Feb 01, 2022
Cloud ComputingIT Strategy

Having learnt from their own migration experiences, IT experts share their advice on how to execute a seamless, successful cloud migration project.

path to cloud computing modern data migration technology picture id1288032037
Credit: iStock

Having learnt from their own migration experiences, IT experts share their advice on how to execute a seamless, successful cloud migration project.

Adoption of cloud computing continues to grow in the Middle East. Accelerated by the pandemic, it’s also been fuelled by large-scale government projects and the availability of an expanding range of data centre and managed service options.

With benefits including lower operating costs, reduced business complexity, scalability and improved security, it’s easy to see its appeal, and the success stories are many, in fields as diverse as supply chain logistics, auto sales and industry.

However, it’s important to keep in mind that executing a successful migration of an enterprise’s systems, such as ERP, HCM (human capital management) or SCM (supply chain management), does come with challenges.

Migrations can often be complex, lengthy and expensive produces that if not done right can potentially result in operational disruptions, reduced employee productivity and unexpected costs.

We spoke to a number of Middle Eastern IT leaders about their experiences migrating to the cloud, and share their advice — along with that of industry analysts — to other CIOs and CTOs looking to make this move themselves.

Know why you’re moving to the cloud

It might feel like everyone’s moving to the cloud, but avoid a knee-jerk response to cloud migration advises Harish Dunakhe, software and cloud research director at analyst firm IDC. Instead take the time to consider why you want to move to the cloud in order to make sure it’s the right decision for your business.

Cloud is the foundation of the software as a service (SaaS) IT model, but whether it’s right for your business depends on the answer to one key question says Umesh Moolchandani, group IT head of the Bin Dasmal Group. “This is ‘how much elasticity do your operations require’?”

The great advantage of cloud is its ability to rapidly scale up or down in response to demand, without any of the hidden costs of service disruptions that often accompanies on-premise infrastructure – especially that which needs to undergo frequent reconfiguration.

“If your organisation is one that requires flexibility and has a huge appetite for computing resources but don’t want the huge CAPEX expenses associated with managing on prem, then cloud is perfect for you,” he says.

Ensure you undertake a risk assessment

With a cloud migration there are lots of points of change relating to networking, connectivity, system access and security. Therefore, it’s imperative to undertake a thorough risk assessment before starting any migration project.

“You need to plan for all contingencies, says Mohamed Saada, chief information technology officer at Amer Group. “I was really playing devil’s advocate with my team, putting on a pessimistic mindset and asking all the scary questions – what if this isn’t working, what if this happens etc.

“Risk assessment is important so you know what to do if any of the negative scenarios play out. Then you don’t panic, as you already know what you need to do.”

Migrate critical systems first

Begin by taking the time to identify your critical systems and shape your cloud space around these, Moolchandani recommends. Then start migrating your larger ‘automation islands’ at the same time.

“These are the ecosystems that work best when integrated together and complement each other’s functionalities. They also tend to be the most resource intensive. After this it’s best to identify ‘grouped’ components of the estate to migrate,” he advises.

“Move all planning applications over at once, for example, rather than try to handle each system or tool independently.”

For companies with greater data maturity — that are already leveraging AI and machine learning, for example — it can be wise to take a ‘data first’ approach­, migrating data first and taking advantage of the abundance of integration tools available to continue using your legacy systems to process it as you complete the journey.

But whatever you do, don’t move systems when they’re at highly critical moments, Moolchandani warns.

“Consider scheduling around events like product releases, quarterly financial reporting and big projects that go live in the production environment.”

Be prepared for the unexpected

It’s always important to have an in-depth cloud migration strategy but however well you plan you have to be prepared for unexpected surprises that could upset your plans.

As Saada found out, even with a full simulation prior to a migration, there are some things that can simply take you by surprise.

“The night we uploaded all our data to the cloud we’d made sure the infrastructure guys had dedicated the company’s whole bandwidth to this upload, as connectivity in Egypt isn’t really the fastest.

“We had an estimation on the time it would take to complete, but the connectivity was much slower than expected. It became a very stressful night as many tasks were delayed as they were dependent on the completion of that upload.

“We were trying to work out what had happened to make the upload so slow, as all the reports showed that we had full bandwidth capacity. Eventually, the next morning, we found out that unbeknown to us, our server had been running a Windows update in the background! In the end we just had to laugh.”

Never forget the end user

When moving any application to the cloud, never forget the user. This is especially true for critical business applications and systems as users will have built up their own ‘best practices’ for getting work done.

“For example, some users export data to Excel because it’s easier to work with. Others open multiple sessions so they can quickly copy and paste between screens,” says Moolchandani.

“Failing to understand what users do, or neglecting to verify they can continue to do so can severely hamper user adoption and engagement.”