通过 Gitlab CI/CD 将 Next JS 应用程序部署到 Kube.netes 集群时，ENV_VARS“未定义”：

Question

I stick to this problem quite long now:

I have a standard NextJS app, which uses environment variables (for client side NEXT_PUBLIC_MY_VAR as well for server side ones MY_OTHER_VAR ).

I use the Gitlab CI-CD AutoDevOps with an tiny custom .gitlab-ci.yml -file (see below).

I have a successful connection to my Kube.netes cluster with Gitlab, and my NextJS app is getting also deployed successfully. (Gitlab/K8s cache cleaned, too)

The only thing where I struggle is to get the process.env.ENV_VARS running. Whatever I tried they are undefined .

I deployed my app manually into the cluster, and mounted a configMap to my deployment ( so a .env.local -file is present at /app/.env.local ONLY THEN the ENV_VARS are set correctly.

So how to set the ENV_VARS when deploying my NextJS app via Gitlab Auto DevOps?

I've tried so far:

• setting ENV_VARS in Gitlab -> Settings -> CI/CD -> Variables

• I added an ARG in my Dockerfile which should pick up the Gitlab CI vars when building the Docker image:

FROM node:alpine AS deps
RUN apk add --no-cache libc6-compat
WORKDIR /app
COPY package.json package-lock.json ./
RUN npm ci 
# ...
# ..

# docker build --build-arg API_URL=http://myApiEndpoint
ARG API_URL
ENV API_URL=$API_URL
RUN npm run build

FROM node:alpine AS runner
WORKDIR /app

ENV NODE_ENV production

# COPY --from=builder /app/next.config.js ./ # TRIED WITH AND WITHOUT
COPY --from=builder /app/public ./public
COPY --from=builder /app/.next ./.next
COPY --from=builder /app/node_modules ./node_modules
COPY --from=builder /app/package.json ./package.json

RUN addgroup -g 1001 -S nodejs
RUN adduser -S nextjs -u 1001
RUN chown -R nextjs:nodejs /app/.next

USER nextjs

EXPOSE 5000

CMD ["npm", "run", "start"]

• I have also added the ENV_VARS also in my next.config.js

module.exports = withSvgr({
    serverRuntimeConfig: {},
    publicRuntimeConfig: {
        NEXT_PUBLIC_API_URL: process.env.NEXT_PUBLIC_API_URL,
        API_URL: process.env.API_URL,
    },
});

• I also added a custom .gitlab-ci.yml file (complete file here):

yaml include: - template: Auto-DevOps.gitlab-ci.yml

# added vars for build
build:
  stage: build
  variables:
    API_URL: $API_URL

This is the error message in my POD

> metashop-frontend.react@1.8.0 start
> next start -p 5000
ready - started server on 0.0.0.0:5000, url: http://localhost:5000
ApiURL alias: undefined #### <<<---- 
API_URL: undefined      #### <<<---- 
TypeError: Only absolute URLs are supported
    at getNodeRequestOptions (/app/node_modules/node-fetch/lib/index.js:1305:9)
    at /app/node_modules/node-fetch/lib/index.js:1410:19
    at new Promise (<anonymous>)
    at fetch (/app/node_modules/node-fetch/lib/index.js:1407:9)
    at Object.getItems (/app/.next/server/pages/_app.js:1194:12)
    at getServerSideProps (/app/.next/server/pages/index.js:2952:55)
    at renderToHTML (/app/node_modules/next/dist/next-server/server/render.js:40:221)
    at processTicksAndRejections (node:internal/process/task_queues:96:5)
    at async /app/node_modules/next/dist/next-server/server/next-server.js:109:97
    at async /app/node_modules/next/dist/next-server/server/next-server.js:102:142

And is the code that the error refers to:

const ApiUrl = process.env.API_URL;

console.log("ApiURL alias:", ApiUrl);
console.log("API_URL:", process.env.API_URL);
console.log("NEXT_PUBLIC_API_URL:", process.env.NEXT_PUBLIC_API_URL);
return fetch(`${ApiUrl}/items.json?${qs.stringify(options)}`).then(
    (response) => response.json()
);

and for completeness (but mostly useless) the tail of the failing job (which seems the normal error when K8s not responding):

Error: release production failed, and has been uninstalled due to atomic being set: timed out waiting for the condition
Uploading artifacts for failed job
00:01
Uploading artifacts...
WARNING: environment_url.txt: no matching files    
WARNING: tiller.log: no matching files             
ERROR: No files to upload                          
Cleaning up file based variables
00:01
ERROR: Job failed: exit code 1

is .env.local the only way to use NextJS ENV_VARS in Kube.netes?

Do I have to customize the Gilab AutoDevOps for this particular (and common) app deployment?

Thank you in advance, any help would be appreciated

Answer 1

If you stick on Auto Devops on GITLAB then YOU HAVE TO set this Gitlab ENV Variable AUTO_DEVOPS_BUILD_IMAGE_EXTRA_ARGS and set it to

--build-arg=NEXT_PUBLIC_API_URL=http://your.domain.com/api

In your Dockerfile you can assign it via

    ARG NEXT_PUBLIC_API_URL
    ENV NEXT_PUBLIC_API_URL=$NEXT_PUBLIC_API_URL

More information here:

https://docs.gitlab.com/ee/topics/autodevops/customize.html#passing-arguments-to-docker-build