From Internet Ops to Banking: Lessons on Data Center Challenges

This article compiles a discussion from the “Efficient Operations” WeChat group series “Sit and Discuss”. The official public account “Efficient Operations” published the original content.

Editor

Dong Wei – Beijing (Compilation & Publication)

Main Participants

Questioner: Xiao Tianguo

Responder: Zhi Jin

Guest Introduction

Xiao Tianguo Director of Operations at Touch Technology, co‑founder of the Efficient Operations technical community. Author of the online column “Best Practices in Efficient Operations”, advocating “Efficient Operations”. Zhi Jin Founder of Hangzhou Yunji Technology, senior operations practitioner promoting “White‑Box Operations”. Former founder of Alipay’s operations team, later led the construction of the private cloud platform for the headquarters of China Construction Bank. Currently focused on developing the best data‑center operating system.

Introduction

“Sit and Discuss” is a rotating Q&A format where, after answering a question, the participant can pose a new question to the next person in the group.

This week is Operations Week, featuring several industry experts exchanging questions.

Key Questions

The questions covered include:

What fundamental differences do you see between traditional financial industry operations and internet industry operations, and what unique experiences have you gained?

You often say, “Make the world free of difficult data‑center operations.” How can this ideal be achieved? Please explain in plain language.

As a founder of Alipay’s operations team, why did you leave, do you regret it, and what were the most frustrating moments at Alipay? Provide two examples.

Q1: What fundamental differences do you see between traditional financial industry operations and internet industry operations, and what unique experiences have you gained?

I worked six years in Alipay’s operations before moving to China Construction Bank in 2011 to build its private cloud and operational tools, a rare transition from internet to a traditional sector. Over the years I have deeply studied “traditional enterprise thinking”.

From the perspective of operational requirements, tasks, and nature, internet and banking operations are similar. However, the methods and approaches differ significantly.

Internet operations serve the business directly; this is the core distinction driven by management philosophy, organizational processes, and business demands of internet enterprises.

Internet operations evolve rapidly due to explosive business growth; companies like BAT treat technology and operations as core competitive advantages and invest heavily. Yet before Alibaba’s 2010 Double‑11 success, the industry viewed Alibaba as a marketing‑driven, non‑technical company.

Traditional enterprises prioritize risk control and stability.

Because traditional businesses have historically stable workloads without sudden spikes, their operational focus is on risk mitigation and stability. Consequently, they rely heavily on legacy hardware and manual management, lacking strong incentives to adopt new technologies or transform operational models.

In short, the fundamental difference lies in differing business demands and corporate goals.

Q2: You often say, “Make the world free of difficult data‑center operations.” How can this ideal be achieved? Explain in plain language.

After interacting with many traditional‑enterprise operators, I found data‑center staff often endure harsh conditions, a result of the nature of the operations field.

Technical Debt Must Be Repaid

IT supports business, and operations support IT; thus operations are the “backend of the backend”. Regardless of company size, operations rarely receive sufficient attention or resources, leading to chronic technical debt.

Even BAT only begins to address debt after reaching a certain scale, often allocating new resources only to the most urgent needs.

Recently I shared an article by Alibaba’s Chen Wei stating that CMDB is the conscience of operations. The internet sector often boasts automation slogans and builds countless systems, yet they cannot keep pace with changing demands, resulting in constant re‑engineering.

Traditional Enterprise Operations Face Growing Pressure

Internet companies already face pressure; traditional enterprises feel it even more:

Internet business drives traditional industry demands. The rise of internet finance forces traditional banks to align with internet technology stacks, dramatically increasing pressure.

Technologies like x86 and virtualization are proliferating in finance. These technologies expand data‑center equipment scale, turning weak demand into strong demand, intensifying operational requirements.

Make the World Free of Difficult Data‑Center Operations

Operations are challenging; only large companies like BAT truly understand this. As industry demand grows, a solution will inevitably emerge.

My slogan, “Make the world free of difficult data‑center operations,” is ambitious and not easy to achieve.

However, “make” does not imply a single product I will create; rather, reflecting on the past decade, I see:

Early best practices like ITIL laid the theoretical foundation for operations.

Mid‑stage pioneers such as Alibaba and Tencent introduced automation best practices.

Today, the entire operations industry—both internet and traditional—is awakening, seeking collective wisdom to solve these challenges, exemplified by the Efficient Operations group.

Thus, “Make the world free of difficult data‑center operations” means leveraging the current era’s momentum, contributing what I can, and rallying the collective intelligence of operators to accelerate this goal.

Q3: As a founder of Alipay’s operations team, why did you leave, do you regret it, and what were the most frustrating moments at Alipay? Provide two examples.

I left Alipay for China Construction Bank, moving my family from Hangzhou to Beijing, a decision many didn’t understand. When I left, the head of Alipay’s tech department said he could understand a move to Tencent or NetEase, but not to a traditional bank.

Remarkably, I was eight months pregnant at the time, and state‑owned enterprises don’t negotiate offers; I didn’t know salary, role, or leadership style before moving.

My vague belief was that internet operations had become homogeneous, while banking operations were still lagging, offering greater opportunities.

Another factor was my strong emotional ties to Alibaba as a product of its training; I felt a narrow rivalry with Tencent and preferred a less conventional path.

I am a risk‑tolerant person; I prepared for worst‑case scenarios and took the leap.

In hindsight, I do not regret the decision. I prepared for the worst and could always return to an internet company if needed.

After arriving, I found the atmosphere and leadership at the bank excellent, coinciding with a decennial IT system overhaul. The bank’s leadership allowed me to apply internet‑style thinking to develop a private cloud and operations platform, exceeding expectations, so I am very grateful for that choice.

The most frustrating moments at Alipay are sensitive and could hurt colleagues, so I will not elaborate further.

Nevertheless, those challenges taught me valuable lessons and contributed significantly to my growth. The world is diverse, and I now understand many perspectives, even if I don’t always agree with them.