Ec2 deploy with Docker #59
Conversation
포크한 레포에 작동 테스틀 위해 연결 스토리북 형태도 세팅함
구조 수정 및 tailwind prefix에 맞게 mapping
이전 패키지 버전에선 빌드가 안 됐음.
새로고침 막고, mutate처리 및 뒤로가기 정상화
todo 개수 0이 아닌 경우에 동작하도록 수정
api 경로에 따라 debounce하도록 수정 및 시간 제한 수정
기존에 수도 클래스로 적용하면 씹히는 경우가 있어서 props를 직접 사용하는 방식으로 수정
5개 넘어야 보이도록
goalId 동일한 것으로 잡도록 수정 dueDate도 계산해서 추가
게스트 모드라면 쿼리 키 다르게 해 refetch 유도 게스트 모드 기록 체크 및 기록이 있다면 온보딩 생략 중복되는 closeMSW함수 제거
게스트 첫 진입 시 onNext실행하도록 수정
도커 허브는 제 개인으로 해두겠습니답
WalkthroughAdds Docker containerization and a production GitHub Actions deploy workflow, introduces a .dockerignore, adjusts token refresh behavior by removing automatic logout/redirect on missing refresh token, and updates onboarding login flow to conditionally advance based on presence of a guest IndexedDB. Changes
Sequence Diagram(s)sequenceDiagram
autonumber
actor Dev as Developer
participant GH as GitHub Actions (main)
participant DH as Docker Hub
participant EC2 as EC2 Host
participant DK as Docker Runtime
Dev->>GH: Push to main
GH->>GH: Checkout, setup
GH->>DH: docker login
GH->>DH: docker build & push tag = ${GITHUB_SHA}
GH->>EC2: SSH (secrets: host/user/key)
EC2->>DK: docker pull ${user}/motimo-fe-production:${SHA}
EC2->>DK: docker stop/rm motimo-fe-production (if exists)
EC2->>DK: docker run -d --name motimo-fe-production -p 3000:3000 --env-file ./production.env IMAGE
note over EC2,DK: Container serves on port 3000
sequenceDiagram
autonumber
actor U as User
participant UI as LoginScreen
participant IDB as IndexedDB
participant Auth as Auth Service
U->>UI: Browse (handleBrowse)
UI->>IDB: Check for guest DB (DB_NAME)
alt Guest DB missing
UI->>Auth: login()
UI->>UI: setIsGuest(true)
UI->>U: onNext()
else Guest DB exists
UI->>Auth: login()
UI->>UI: setIsGuest(true)
UI->>U: Stay (no onNext)
end
sequenceDiagram
autonumber
participant App as App
participant Token as tokenHandler
participant API as API Server
App->>Token: Request with expired access token
alt Has refresh token
Token->>API: Reissue access/refresh
API-->>Token: New tokens
Token-->>App: Proceed with updated tokens
else No refresh token
Token-->>App: Throw "no refresh token" (no logout/redirect)
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Poem
Tip 🔌 Remote MCP (Model Context Protocol) integration is now available!Pro plan users can now connect to remote MCP servers from the Integrations page. Connect with popular remote MCPs such as Notion and Linear to add more context to your reviews and chats. ✨ Finishing Touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR/Issue comments)Type Other keywords and placeholders
CodeRabbit Configuration File (
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (2)
.dockerignore (1)
1-4: Missing trailing newline at end of file.This is a minor formatting issue, but it's generally good practice to end files with a newline character.
Add a trailing newline at the end of the file:
.next .git node_modules .gitignore +Dockerfile (1)
10-11: Consider using npm ci for production builds.Using
npm ciinstead ofnpm installprovides faster, reliable, reproducible builds by installing directly from package-lock.json without modifying it.Replace
npm installwithnpm cifor production builds:-# 의존성을 설치합니다. -RUN npm install +# 의존성을 설치합니다. +RUN npm ci --only=production
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (5)
.dockerignore(1 hunks).github/workflows/production-deploy.yml(1 hunks)Dockerfile(1 hunks)api/service.ts(1 hunks)app/onboarding/_components/LoginScreen.tsx(1 hunks)
🧰 Additional context used
🪛 actionlint (1.7.7)
.github/workflows/production-deploy.yml
16-16: the runner of "actions/checkout@v3" action is too old to run on GitHub Actions. update the action's version to fix this issue
(action)
20-20: the runner of "docker/login-action@v2" action is too old to run on GitHub Actions. update the action's version to fix this issue
(action)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: chromatic-test
🔇 Additional comments (5)
api/service.ts (1)
82-84: Verify removal of logout + redirect in 401/no-refresh-token path.
Commenting out api.authController.logout() / window.location.href="/" in api/service.ts (lines 82–84) changes the UX: a missing refresh token now throws instead of auto-logging-out/redirecting. I found other explicit logout usages (components/mypage/MyPage.tsx, app/mypage/edit/page.tsx). Confirm this is intentional and that global or caller error handling will reliably route users to login (or restore the logout+redirect).app/onboarding/_components/LoginScreen.tsx (1)
181-188: Conditional onboarding flow looks correct.The logic to check for existing guest database and conditionally call
onNext()prevents users with existing guest data from repeating the onboarding process. This improves user experience by maintaining state across sessions.The implementation correctly uses IndexedDB to detect existing guest data and sets the appropriate completion flag.
Dockerfile (2)
18-21: Environment variable configuration looks good.The ARG and ENV pattern correctly allows build-time injection of the frontend base URL, which is essential for containerized deployments across different environments.
7-14: Optimize Docker layer caching.The current approach copies all files before installing dependencies, which breaks Docker layer caching. When source code changes, dependency installation will re-run unnecessarily.
Reorder the COPY commands to optimize layer caching:
# package.json과 package-lock.json (또는 yarn.lock)을 복사합니다. COPY package*.json ./ -# 의존성을 설치합니다. -RUN npm install - -# 프로젝트의 모든 파일을 작업 디렉토리로 복사합니다. -COPY . . +# 의존성을 설치합니다. +RUN npm ci --only=production + +# 프로젝트의 모든 파일을 작업 디렉토리로 복사합니다. +COPY . .Likely an incorrect or invalid review comment.
.github/workflows/production-deploy.yml (1)
49-49: Verify production.env file exists on EC2.The deployment assumes a
production.envfile exists in the home directory of the EC2 instance. If this file doesn't exist, the container will fail to start.Ensure that the
production.envfile is properly created on the EC2 instance before deployment, or add error handling:# 새 컨테이너 실행 - docker run -d --name motimo-fe-production -p 3000:3000 --env-file ./production.env ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} + # production.env 파일 존재 확인 + if [ ! -f ./production.env ]; then + echo "Error: production.env file not found" + exit 1 + fi + + docker run -d --name motimo-fe-production -p 3000:3000 --env-file ./production.env ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }}
|
|
||
| steps: | ||
| - name: Checkout source code | ||
| uses: actions/checkout@v3 |
There was a problem hiding this comment.
Update GitHub Actions to newer versions.
The static analysis correctly identified that the actions versions are outdated and may not work on current GitHub runners.
Update the actions to newer versions:
- name: Checkout source code
- uses: actions/checkout@v3
+ uses: actions/checkout@v4
# 1. Docker Hub에 로그인 (ECR 로그인 대신)
- name: Login to Docker Hub
- uses: docker/login-action@v2
+ uses: docker/login-action@v3Also applies to: 20-20
🧰 Tools
🪛 actionlint (1.7.7)
16-16: the runner of "actions/checkout@v3" action is too old to run on GitHub Actions. update the action's version to fix this issue
(action)
🤖 Prompt for AI Agents
In .github/workflows/production-deploy.yml around lines 16 and 20, the
referenced GitHub Action versions are outdated; update the actions to their
current stable releases (e.g., bump actions/checkout@v3 to actions/checkout@v4
and similarly upgrade the action used at line 20 to its latest major version),
ensure the workflow syntax and inputs remain compatible after the upgrade, and
run a workflow test to confirm no breaking changes.
| docker build -t ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} . | ||
| docker push ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} |
There was a problem hiding this comment.
Add build argument for environment variable.
The Docker build command doesn't pass the required build argument for NEXT_PUBLIC_FRONTEND_BASE_URL, which will result in an undefined environment variable.
Add the build argument to the Docker build command:
- name: Build and push to Docker Hub
run: |
- docker build -t ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} .
+ docker build --build-arg NEXT_PUBLIC_FRONTEND_BASE_URL=${{ secrets.NEXT_PUBLIC_FRONTEND_BASE_URL }} -t ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} .
docker push ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }}📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| docker build -t ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} . | |
| docker push ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} | |
| - name: Build and push to Docker Hub | |
| run: | | |
| docker build --build-arg NEXT_PUBLIC_FRONTEND_BASE_URL=${{ secrets.NEXT_PUBLIC_FRONTEND_BASE_URL }} -t ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} . | |
| docker push ${{ secrets.DOCKERHUB_USERNAME }}/motimo-fe-production:${{ github.sha }} |
🤖 Prompt for AI Agents
In .github/workflows/production-deploy.yml around lines 28 to 29, the docker
build command doesn't pass the NEXT_PUBLIC_FRONTEND_BASE_URL build argument,
causing that env var to be undefined; update the docker build invocation to
include a build-arg flag passing the secret (e.g. --build-arg
NEXT_PUBLIC_FRONTEND_BASE_URL=${{ secrets.NEXT_PUBLIC_FRONTEND_BASE_URL }}) so
the value is available at image build time before pushing.
| if [ $(docker ps -q -f name=motimo-fe-production) ]; then | ||
| docker stop motimo-fe-production | ||
| docker rm motimo-fe-production | ||
| fi |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Improve container cleanup robustness.
The current container cleanup approach may fail if containers exist but are already stopped, potentially causing deployment failures.
Make the cleanup more robust by handling different container states:
# 기존 컨테이너 중지 및 삭제
- if [ $(docker ps -q -f name=motimo-fe-production) ]; then
- docker stop motimo-fe-production
- docker rm motimo-fe-production
- fi
+ # 실행 중인 컨테이너가 있다면 중지
+ if [ $(docker ps -q -f name=motimo-fe-production) ]; then
+ docker stop motimo-fe-production
+ fi
+
+ # 존재하는 컨테이너가 있다면 삭제
+ if [ $(docker ps -aq -f name=motimo-fe-production) ]; then
+ docker rm motimo-fe-production
+ fi📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| if [ $(docker ps -q -f name=motimo-fe-production) ]; then | |
| docker stop motimo-fe-production | |
| docker rm motimo-fe-production | |
| fi | |
| # 기존 컨테이너 중지 및 삭제 | |
| # 실행 중인 컨테이너가 있다면 중지 | |
| if [ $(docker ps -q -f name=motimo-fe-production) ]; then | |
| docker stop motimo-fe-production | |
| fi | |
| # 존재하는 컨테이너가 있다면 삭제 | |
| if [ $(docker ps -aq -f name=motimo-fe-production) ]; then | |
| docker rm motimo-fe-production | |
| fi |
🤖 Prompt for AI Agents
.github/workflows/production-deploy.yml around lines 43 to 46: the current
cleanup uses `docker ps -q -f name=motimo-fe-production` and only stops/removes
if running, which can fail when a container exists but is stopped; replace the
logic to query all containers by name (`docker ps -a -q -f
name=motimo-fe-production`) and then force-remove any found container (e.g., use
`docker rm -f <id>`), or if you prefer two-step: if container id exists, attempt
`docker stop <id>` (ignore errors if already stopped) then `docker rm <id>`;
ensure commands run safely in CI (handle empty id) so cleanup succeeds
regardless of container state.
1 participant
배포
도커 사용해서 ec2에 배포하는 방식을 채택했습니다.
ec2 프리티어라서 인스턴스 내부에서 빌드가 힘들어서요
github action통해 cd 자동화 했습니다.
버그 수정
Summary by CodeRabbit
Bug Fixes
Chores