Bring your towel and join one of the fastest growing data communities. Welcome to our second-generation open source orchestration platform, a completely rethought approach to dataflow automation.

Prefect Community

Capture d’écran 2024-10-12 à 11.46.50.png

Sry to re-ask, but I wanna understand the meaning of each parameter for `Concurrency` :

I have multiple API key, so TMDB have rate limit of 50 requests / seconde so I just dont wanna hit this rate limit. But what `Concurrency Limit`, `Active Slots` are made for ?

hi <@U06CLKHAW4W> - the docs on this are <https://docs.prefect.io/3.0/develop/global-concurrency-limits#slot-decay|here>!

slot decay is just a way to set a time-based strategy where concurrency slots are released after the number of seconds you specify

But if I just wanna protect the rate limit, like dont go above 50 requests per seconde, what should I set ? Because I dont care how many concurrency there is, just wanna slow down when rate limit is reached :sweat_smile:

Because if I well understood, if I increase the `Slot Decay per Second` its gonna be slower right ?

Well Ive wrong i guess :
```To configure slot decay, set the slot_decay_per_second parameter when creating or updating a concurrency limit. This value determines how quickly slots refresh:

A higher value (for example, 5.0) means slots refresh quickly. Tasks can run more frequently, but with short pauses between them.
A lower value (for example, 0.1) means slots refresh slowly. Tasks run less frequently, with longer pauses between them.```
Mhh Im kinda lost here :sob:

&gt; Because I dont care how many concurrency there is, just wanna slow down when rate limit is reached :sweat_smile:
then Im not sure you need to worry about slot decay at all - just keep it at zero and use the `rate_limit` ctx manager and it sounds like what you describe is the default behavior

But if I dont set a limit how can I know if the rate limit is reached. Sry Im kinda new to python I lack some notions

<@U06CLKHAW4W> You want 50 RPS, which means you will need to allow 50 slots every second which is 50 concurrency limit and slot decay of 50