Cloud Function connection exhaustion
This Cloud Function (Python) queries Postgres on each HTTP request. The AI generated it and it works in local testing. Under traffic the database starts rejecting connections with 'too many clients already'. What's wrong and how do you fix it?
Implement
peak_db_connections(instances: int, requests_per_instance: int, close_after_use: bool, pooler_limit: int) → intExamples
in
[200,5,false,0]out1000in
[200,5,true,0]out200in
[200,5,false,20]out20What a strong answer looks like
Treat the AI’s output as a draft to verify, not an answer to trust. Name the specific flaw and the input that triggers it, say how you’d catch it (tests, edge cases, reading critically), and how you’d re-prompt or decompose to get it right.
0:00 of about 16 min
Vibe & agentic: describe the solution in plain language (or narrate it) and the coach grades your approach.
Which questions mattered is sealed until you submit. Telling you now would just be handing over the edge cases.
Run or narrate your approach, then ask the coach.