How Lotte Shopping eCommerce Optimized MSA Monitoring Using WhaTap
Lotte Shopping eCommerce, one of WhaTap’s customers, operates Lotte ON, a platform that integrates online and offline to provide a convenient customer experience. Lotte ON, a giant retailer created by combining seven business divisions online and offline, leveraging its strengths as a traditional retail powerhouse. In December 2022 and January 2023, LotteON exceeded 2 million monthly active users (MAUs) for the first time, and its average daily traffic has recorded 330,000.
Lotte ON switched to WhaTap monitoring after having unpleasant experience with other monitoring services. To delve deeper into why Lotte Shopping eCommerce, an enterprise company, chose WhaTap and how they are utilizing it, we conducted an interview with its employee, Jeong Sung-min.
Introduction: Introducing client and infrastructure
Please provide us with a brief introduction to LOTTE ON and its services.
LOTTE ON is an integrated e-commerce platform that brings together Lotte's representative shopping malls, making them easily accessible to users. It provides a convenient customer experience by offering access to Lotte Home Shopping, Hi-Mart, and Super Fresh on a single platform. We are dedicated to delivering fresh stories to customers by seamlessly integrating online and offline services.
LOTTE companies offer a wide range of services. What kind of environment have you been working in?
We have embarked on various digital journeys to provide our services. Starting with the establishment of Lotte Internet Department Store in 1996 in an on-premise environment, we transitioned to a cloud environment in 2015-16 and launched services for Nike and Uniqlo. Subsequently, we developed Lotte Internet Duty Free as a MSA in a cloud environment. In 2018, we constructed Ellotte in a cloud-native environment, and finally, we built the LotteON service.
Challenge: How you started using WhaTap
I'm curious about your background and the reasons behind choosing the WhaTap monitoring service.
The LOTTE ON service is hosted in the AWS cloud with an EKS and MSA configuration. Initially, LOTTE ON's architecture presented a significant challenge in terms of monitoring. Given the complexity of the architecture, we needed a solution that could effectively monitor Kubernetes. At the time, there were limited options for Kubernetes monitoring services, and we also had concerns about efficiently delivering the MSA-divided area to practitioners. While we considered other offshore monitoring products, we ultimately opted for WhaTap.
The primary factor in choosing WhaTap was its intuitive and user-friendly UI. We believed that APM should be accessible and comprehensible to anyone, not just developers. The intuitive dashboard and familiar UI of WhaTap proved to be incredibly valuable in swiftly identifying and addressing issues.
We're curious to hear how you've been using WhaTap since its introduction.
There are three features that we utilize extensively. First is the dashboard status. We monitor a dashboard that graphs various data, such as the number of payments, orders, etc., which are key business indicators for LOTTE ON services. This helps LOTTE ON respond quickly to other affected business services. For example, if an order fails due to an issue with a third-party credit card company, we can promptly address the problem by adjusting the payment method.
The second feature is the flexible alert feature. Since our service is divided into MSAs and each person is in charge of specific areas, we have separate Slack channels. We can set thresholds for each MSA service issue in each channel, allowing each person to receive the appropriate notification.
The third and final feature is the statistics/reports feature, which enables us to view the number of exceptions for each MSA on a weekly basis. We share this statistical information with each representative to identify and address areas that require action. This feature has been particularly helpful in analyzing the causes of problems, as it allows us to identify the peak times for exceptions during failures.
Please tell us about an experience you had while using WhaTap that you were able to resolve.
I can share two experiences in this regard. The first one involved an issue with a point-earning event. We were offering points to customers who filled out a survey on a specific product, but the event was mistakenly applied to all products. As a result, the event quickly gained attention on various internet community sites, leading to a surge in traffic. This caused a significant increase in CPU usage for the POD, resulting in transaction processing delays.
In this situation, WhaTap's EKS POD monitoring dashboard proved to be highly effective. It allowed us to monitor the resource status of each POD container in real time, providing an intuitive monitoring experience. The ability to set customized threshold settings was particularly helpful, as it triggered alarms based on specific conditions and immediately alerted the relevant staff. Additionally, utilizing statistical indicators enabled us to identify problematic URLs during that time, enabling us to promptly take measures to control them and prevent future issues.
The second issue occurred during a large-scale event that we hold twice a year, spanning approximately a week, where we distribute a large number of coupons to our customers. The sheer volume of coupons created a heavy load on our logic for applying maximum discounts at checkout, resulting in an Out-of-Memory (OOM) issue in the discounting POD. For this issue, we utilized WhaTap's Heap monitoring feature. By comparing the graph showing the time of the issue to normal, we discovered a significant increase in the number of SQL patches. This insight allowed us to address the problem by implementing preventive measures to avoid similar issues in the future. Additionally, we configured heap memory thresholds for each container, enabling us to take immediate action at the first sign of OOM, which has proven beneficial thus far.
Management: WhaTap customer support and future plans
PWhat are your plans for WhaTap in the future?
In addition to the examples mentioned above, we have encountered various challenges since transitioning to an MSA (Microservices Architecture). This shift has resulted in multiple services being interconnected within a single transaction, leading to dispersed potential risks. Rather than solely detecting issues through an APM (Application Performance Monitoring) dashboard, we increasingly require performing diverse analyses within this complex architecture. Our mission is to progress from simple monitoring to observability, and we are diligently leveraging all the metrics collected in WhaTap to gain proactive insights and assist LOTTEON in achieving observability. We are still collaborating with WhaTap engineers to utilize additional metrics effectively.
Tell us why we should use WhaTap monitoring solution.
Renowned management scholar Peter Drucker once said, "If you can't measure it, you can't manage it." To transition from mere monitoring to comprehensive quality management and attain observability, it is crucial to collect extensive data and leverage it effectively. WhaTap's monitoring solution is indispensable for achieving this observability by facilitating data collection and utilization at the right time.