Kakao Reveals Development Journey of Mobile App CI/CD Platform 'MobiL'
Key point
Developed since 2013, MobiL supports automated builds based on Jenkins and MacPro, along with deployment via an internal-only app center.
Details
Kakao has revealed the development background and architecture of MobiL, a CI/CD platform developed in-house to automate the build and deployment processes of mobile apps. In the early stages of development in 2013, specialized CI/CD tools for mobile apps were scarce, and manual deployments and inefficient communication were widespread. MobiL aims to automate every step of the release process so that developers and QA can focus on their core tasks.
Core Features and Principles
MobiL's overarching principle is Easy, designed to allow immediate build execution with minimal information input. It offers three build configuration types—CI, PRERELEASE, and RELEASE—and six build execution methods, including Smart Commit and Scheduled Build. In particular, the Manual Build feature allows anyone to execute a build at any desired time, even if they are not the designated person, thereby supporting Agile Testing.
Architecture and Deployment Method
MobiL consists of a web service and an internal-only app called Kakao App Center. When a build is requested, Jenkins creates a Job, but the actual build is performed on a MacPro registered as a Slave Machine. Build artifacts are stored in the distributed storage system KAGE and deployed via OTA through the Kakao App Center. Additionally, it integrates with the crash analysis system MOCA and the issue tracking system JIRA, allowing users to grasp build information and issues at a glance.
Future Plans
As the next step for MobiL, Kakao is pursuing UI Test Automation on real devices. The goal is to establish an environment where automated tests can be performed and results received across various devices and versions, while continuously exploring ways to enhance usability.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.