Oliveyoung Payment Story Part 2

As service growth increased the need to grasp the status of the order and payment system in real time, Oliveyoung introduced Datadog to build a monitoring system. Previously, the team relied on text messages sent when orders failed, making it difficult to immediately determine the cause of an incident (online mall vs. PG company), and since these were mixed together with normal failure cases, response was delayed.

For effective monitoring, the following items were implemented on the dashboard.

• Step-by-step failure detection: the entire process from order form entry, to payment window invocation, authentication/approval, and order completion
• Application of various criteria: by channel (MW, APP, PC), by OS/browser, and by payment method (credit card, Naver Pay, Kakao Pay, etc.)

During implementation, logs were designed to be left in a KEY=VALUE pattern so they could be easily searched in Datadog. This made it possible to intuitively look up payment methods or progress status, and by integrating with Slack, an environment was established to receive real-time alerts when an incident occurs and respond quickly.

Currently, to improve the log analysis method, the team is working on further enhancements using Grok Parser to make it easier to extract attributes from logs.