My Next.js app deploys successfully in preview and builds fine locally using a .env.
However, the Vercel production deploy (triggered via GitLab CI) fails with multiple
FetchError: request to https://graphql.contentful.com/... failed, reason: connect ETIMEDOUT
messages during static page generation.
The build logs show production uses
CMS_CONTENTFUL_ENV_ID=master
which points to a much larger Contentful environment.
It seems like the connection to Contentful times out while pre-rendering thousands of pages, even though all environment variables are correctly loaded via vercel pull.
Has anyone encountered this kind of Contentful timeout in Vercel’s production builds?
What’s the best approach — increase the Next.js static generation timeout, optimize queries, or handle this differently in CI/CD?
Or will simply redeploying work?