Concerns about s3-reverse-proxy Performance

[nphamvn]

The s3-reverse-proxy works fine, but I'm concerned about the server's ability to handle the high cost of HTTP requests. Specifically, the server receives the origin HTTP request and then sends another HTTP request to S3. Given the potential volume of requests—n (app instances) * m (users per app)—I’m unsure if the server can efficiently handle such a large number of requests.

[harshsuiiii]

The s3-reverse-proxy works fine, but I'm concerned about the server's ability to handle the high cost of HTTP requests. Specifically, the server receives the origin HTTP request and then sends another HTTP request to S3. Given the potential volume of requests—n (app instances) * m (users per app)—I’m unsure if the server can efficiently handle such a large number of requests.

ANS:

Use CloudFront in front of S3 (Best & Cheapest)

CloudFront is AWS’s CDN.

You can directly attach S3 as the origin (private bucket + signed URLs or OAC).

Benefits:

Global edge caching (less load on S3 + fewer direct requests).

Cheaper per GB bandwidth than S3 public requests.

Offloads HTTPS termination (no need for ECS to sit in middle).