502 Bad Gateway error on front end, missing variable errors on server container

[Uncanny5180]

Hi All, somewhere in the last week my sparkyfitness install stopped working. No changes on my end that I am aware of, and I dont even think I tried to pull a newer version... but I could be wrong on that.

When you go to the front end, it looks like a proper login screen, but a second or two later, you get the red error box, saying 502 bad gateway:

Front end console logs:

172.26.0.15 - - [04/Nov/2025:15:12:49 +0000] "GET /locales/en/translation.json HTTP/1.1" 200 5198 "https://fitness.tld.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36"
172.26.0.15 - - [04/Nov/2025:15:12:52 +0000] "GET /api/version/current HTTP/1.1" 502 559 "https://fitness.tld.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36"
2025/11/04 15:12:52 [error] 15#15: *67 connect() failed (113: Host is unreachable) while connecting to upstream, client: 172.26.0.15, server: localhost, request: "GET /api/version/current HTTP/1.1", upstream: "http://172.22.0.3:3010/version/current", host: "fitness.tld.com", referrer: "https://fitness.tld.com/"
2025/11/04 15:12:52 [error] 13#13: *74 connect() failed (113: Host is unreachable) while connecting to upstream, client: 172.26.0.15, server: localhost, request: "GET /openid/api/me HTTP/1.1", upstream: "http://172.22.0.3:3010/openid/api/me", host: "fitness.tld.com", referrer: "https://fitness.tld.com/"
172.26.0.15 - - [04/Nov/2025:15:12:52 +0000] "GET /openid/api/me HTTP/1.1" 502 559 "https://fitness.tld.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36"
127.0.0.1 - - [04/Nov/2025:15:12:54 +0000] "GET / HTTP/1.1" 200 2224 "-" "curl/8.14.1"
2025/11/04 15:12:55 [error] 15#15: *67 connect() failed (113: Host is unreachable) while connecting to upstream, client: 172.26.0.15, server: localhost, request: "GET /api/auth/settings HTTP/1.1", upstream: "http://172.22.0.3:3010/auth/settings", host: "fitness.tld.com", referrer: "https://fitness.tld.com/"
172.26.0.15 - - [04/Nov/2025:15:12:55 +0000] "GET /api/auth/settings HTTP/1.1" 502 559 "https://fitness.tld.com/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/142.0.0.0 Safari/537.36"
127.0.0.1 - - [04/Nov/2025:15:12:59 +0000] "GET / HTTP/1.1" 200 2224 "-" "curl/8.14.1"

From the backend server container complaining of missing variables:

[[email protected]] injecting env (0) from ../.env – [tip] encrypt with dotenvx: https://dotenvx.com
[2025-11-04T14:55:25.942Z] [ERROR] FATAL: Missing required environment variables: SPARKY_FITNESS_APP_DB_USER, SPARKY_FITNESS_APP_DB_PASSWORD. Please check your .env file.
/app/SparkyFitnessServer/utils/preflightChecks.js:21
    throw new Error(errorMessage);
    ^
Error: FATAL: Missing required environment variables: SPARKY_FITNESS_APP_DB_USER, SPARKY_FITNESS_APP_DB_PASSWORD. Please check your .env file.

The variables are and have been present... I have no idea why its so unhappy. I even put the actual data (db name, user and pass) into the docker compose file, instead of it referring to the .env, and no change. Any suggestions!?

[CodeWithCJ]

As the error shows, add these two variables. This is new and not to be confused with previous credentials.

SPARKY_FITNESS_APP_DB_USER, SPARKY_FITNESS_APP_DB_PASSWORD

I'll look for a way to avoid this additional APP DB USER. But currently this is needed for better security for supporting multiple users

[Uncanny5180]

got it... thanks, apparently I was just not 'seeing' the APP_ bit in those variables... I got my docker files updated and she's alive again.