Llm Gateway Cost And Caching easy
Your gateway applies one 5-second timeout to all LLM calls. Classification calls work fine, but long streamed answers are frequently cut off. What is the best fix?
- A Raise the timeout to 120 seconds for all calls so no request is ever cut off early.
- B Keep 5 seconds everywhere and retry streamed calls until one finishes in time.
- C Set timeouts by call type: long for streaming, short for classification.
- D Remove timeouts entirely and rely on the provider to close finished connections.
Answer: C. Set timeouts by call type: long for streaming, short for classification.
**Timeouts should match the call type**: streamed generation legitimately runs for many seconds, while a stuck classification call should fail fast so it can be retried or rerouted. A single long timeout fixes the cut-offs but lets hung short calls tie up resources. Retrying long streams with a short timeout just pays repeatedly for output that is never delivered.
One system design question a day, with spaced repetition so the ones you miss come back.
Get the appRelated questions
- Your primary LLM provider is fully down for an hour, and no tested fallback exists for your agent feature. The product also has an FAQ answer feature backed by a semantic cache. What is the best degradation mode?
- A multi-tenant B2B support assistant answers questions using each customer's private knowledge base. To raise the semantic cache hit rate from 22% to 35%, an engineer proposes one shared cache across all tenants. What is the main problem?
- During a provider brownout returning 429 and 503 errors, the client SDK retries 3 times and the gateway also retries 3 times, both immediately. Provider load spikes, and the bill grows from partially streamed responses. What should you change?
- Finance asks for LLM spend broken down by tenant and by product feature. Today the gateway only sees the total on the provider's monthly invoice. What should you do?
- A retail assistant uses a semantic cache. Weeks after a store changed its holiday hours, users still receive the old hours from cached answers. What is the right fix?
- A feature sends 1,500 input tokens and receives 600 output tokens per request. Output tokens cost about 5x input tokens. Which single change cuts cost the most?