Skip to content

0006-optimize-images-with-imagekit.md

Latest commit

 

History

History
31 lines (19 loc) · 1.59 KB

0006-optimize-images-with-imagekit.md

File metadata and controls

31 lines (19 loc) · 1.59 KB
Raw

ADR-0006: Offload Media Optimization to ImageKit CDN

Status: Accepted
Date: -
Author: Mehdi Maleki

Context

Cinemark is a highly visual application. Movie posters must be served in multiple sizes, optimized for modern formats (WebP/AVIF), and delivered with low latency globally.

Decision

We will use ImageKit as our media CDN and transformation engine, integrating it via the imagekitio-react SDK and Next.js Image component fallbacks.

Alternatives Considered

  • Next.js Built-in Image Optimization: Rejected. While convenient, it consumes Vercel's bandwidth and optimization limits. It also lacks advanced transformation features (like dynamic watermarking or complex cropping) without custom code.
  • AWS S3 + CloudFront + Lambda: Rejected. Building a custom image transformation pipeline using Lambda@Edge is highly complex, expensive, and difficult to maintain for a solo developer.

Consequences

Positive

  • Zero Compute Overhead: Image transformations happen at the CDN edge, saving CPU cycles on our Vercel serverless functions.
  • Bandwidth Savings: Automatic WebP/AVIF conversion significantly reduces payload sizes, improving Core Web Vitals (LCP/CLS).
  • Bandwidth Offloading: Offloads image traffic from Vercel's hosting limits to ImageKit's specialized infrastructure.

Negative & Risks

  • Vendor Lock-in: Our image URLs and transformation parameters are tied to ImageKit's syntax.
  • Mitigation: We abstract the ImageKit logic inside a reusable <MovieCover /> component. If we ever need to migrate, we only update this single component.