Expansion Leads to a New Solution for Mission-Critical Infrastructure

When it comes to mission-critical IT environments, you want three things: reliability, availability, and serviceability (or RAS). That means choosing an infrastructure that is easy to configure and maintain but powerful enough to run the applications that drive your business. However, striking the right balance between simplicity and power can be challenging when the demands on your IT are constantly evolving to meet the needs of thousands of internal and external users scattered across a wide geographic area.

Yapı Kredi is Turkey’s first retail-focused private bank with a nationwide presence and the third largest private bank in the country. Founded in 1944, we offer personal subject matter experts and enterprise corporate and commercial banking services. 

I’ve been with Yapı Kredi since 2010 and am responsible for Yapi Kredi’s System Design and Planning team. My team's responsibilities are Linux and UNIX Systems, x86 infrastructure, on-prem cloud infrastructure, and the infrastructure for 3,000 virtual servers.

We have done many new projects over the past 11 years. Our new project is migrating our RISC-based SPARC servers to x86-based technology bare metal or virtual machines. In every case, we’ve adopted new technology to do more with less and increase scalability and flexibility. We have also upped our infrastructure’s RAS by automating system management and disaster recovery tools, thereby ensuring the continuity of mission-critical workloads and minimizing downtime.

Adopting New Tools to Support Our Growth

Given the bank’s steady growth trajectory, we constantly add new services for our customers and new tools for our internal resources. We recently launched a Smart Assistant application that pushes various alerts to individual and institutional customers and investors, including account statements, payment reminders, stock recommendations, buy/sell alarms, and deposit renewals. We also adopted the RStudio data science platform. 

We have also begun to use HPE Synergy, a composable, software-defined infrastructure for server virtualization environments. HPE Synergy enables us to compose pools of physical and virtual compute, storage, and fabric resources into any configuration for any workload under a unified API. We’ve used HPE Synergy to spin up reliable, available, and serviceable virtual servers—in other words, increasing virtualization across our organization.

The combination of highly configurable HPE BladeSystem servers and HPE Synergy’s single-pane-of-glass infrastructure tools gave us the flexibility we needed to meet our inward- and customer-facing IT needs. But as our business continued to grow throughout the global pandemic, we had to scale our infrastructure faster than expected, which meant looking for even more performance, flexibility, and scalability.

Choosing a Trusted Partner to Refresh Our Infrastructure

As you can tell, we’ve enjoyed a long-term business relationship with HPE. The company is solution-oriented and customer-centric, with high-quality support and a diverse range of products. 

As Infrastructure, Network and Operations Deputy General Manager Hakkı Erdoğan says, the most important reasons to choose HPE solutions are: their solution-oriented and customer-centered approach; their high-quality support; and their diversity of solutions. Given our history, we trusted both HPE and their products. We knew they had to have something to meet our growing needs.

They did—the best solution for our digital transformation is the combination of HPE Synergy, HPE Apollo servers, and HPE Superdome Flex 280 servers. We now run high-critical applications and digital platforms on these HPE Servers.

HPE Superdome Flex 280 Servers Offer Power and Flexibility

We built a new environment using HPE Superdome Flex 280 servers. This line of servers features third-generation Intel Xeon scalable processors and a modular architecture that can scale to meet future growth. With its modular building blocks, the Superdome Flex 280 can scale from 16 cores to 224 cores and 64 GB to a staggering 24 TB of shared memory. Its basic configuration can handle today’s complex banking applications and leave us enough room to grow as our product offerings and our internal systems evolve. 

HPE Superdome Flex 280 gives Yapı Kredi’s IT team the freedom to explore our options when embarking on new projects. We no longer worry that we’ll run out of capacity because we can add processors and memory as needed. But this is only the beginning. Over the next five years, we’ll be phasing out the last of our legacy Unix-based servers. Going forward, our production environment and banking applications will run exclusively on x86 Linux servers. 

Further Reducing Complexity

We initially chose blade servers to decrease the cost of cooling and heating, reduce floor space in our data center, and increase productivity. Over time, our team built a wealth of experience with blade technology.

Using a single operating system will further reduce complexity and streamline our operations. HPE Superdome Flex 280 servers are easy to install and maintain, like blade servers. There’s less cabling, and the scalability of features is astonishing. We can add processors and memory modules as needed instead of ripping out old servers to install new ones. Our HPE Superdome Flex 280 servers also integrate seamlessly with the company’s other platforms, including the HPE Apollo servers we use for our AI data applications.

We have also begun to use HPE OneView to automate and simplify lifecycle management and enable rapid DevOps on HPE Synergy. The combination of HPE OneView, HPE Synergy, and HPE Superdome Flex 280 servers allows us to add capacity in minutes instead of hours, automate repetitive infrastructure management tasks, and install updates with a single click. 

We manage more than 3,000 virtual servers. Because our technology has gotten more scalable, we haven’t had to increase our staff to manage our infrastructure as it grows. And because our technology has also gotten more intelligent, my team has more time to focus on engineering instead of spending their days managing our infrastructure and dealing with configuration issues. We are happier, more productive, and experience less stress because we can work on projects that impact the bank’s business outcomes rather than troubleshooting our servers and network.

There’s No Room for Error in Banking

There is no room for error in banking. A millisecond of lag can affect the value of a stock or currency trade. A server or network outage may prevent a customer from making a withdrawal or paying for an in-store purchase with their debit or credit card.

HPE solutions are reliable, stable and address all of those concerns. HPE infrastructure meets all our RAS criteria, and their team has demonstrated incredible support over the years. We rely on HPE to be highly available and easily serviced, ensuring 24/7 availability of our environment for internal users and customers alike.