This looks like a transient DNS timeout inside the Netlify build environment rather than a true DNS misconfiguration, especially since it works locally and the Heroku app resolves globally. EAI_AGAIN typically indicates a temporary DNS resolution failure, often caused by network flakiness, IPv6 resolution issues, or high concurrency during build-time data fetching. I’d recommend first adding retry logic with exponential backoff to your Gatsby/Strapi fetch requests to handle intermittent failures gracefully. Additionally, set the environment variable NODE_OPTIONS=--dns-result-order=ipv4first in Netlify to avoid potential IPv6 resolution instability on CI runners. If the issue persists, consider reducing parallel requests during sourcing or placing a caching layer (e.g., CDN) in front of the Heroku app to stabilize responses during builds. If multiple projects are affected, it’s also worth sharing failing build IDs with Netlify support to check for regional resolver issues on specific build runners.