Bounded queue under a token limit
Difficulty 3 of 5
Each upload needs a summary: 4,000 input and 400 output tokens. The provider allows 600,000 tokens/min and rejects the rest. From 0:20 to 1:00, uploads jump from 1/s to 8/s. Keep upload p99 below 1 s, jobs under 60 s old, rejections at most 35 percent, and lose no job.
Buffer 1,000 jobs · last job waits 1.0 s · 17% turned away
Goals and limits
- p99 latency
- under 1 s
- Queue lag
- under 60 s
- Error rate
- under 35%
- Lost writes
- none
- Monthly cost
- under 40 USD/mo
- Parts you can add
- Managed queue, Message broker, Worker
- Part limit
- 2
- Traffic
- 1 req/s
Concepts
How players did
Player results load from the server.