Hi @tomasw!
Thank you for such a quick response.
Our query is pretty optimized…as it only uses 10 credits per second (and we’re refreshed with 15 credits every second, so we’re within the allotted usage). We use a queueing service (AWS SQS) with our serverless ELT process to queue up larger credit consuming queries as to self throttle these requests to staying within the allotted credits.
The issue we are running into is the credit limit per hour… which, as I understand, is set at 5,000. This is waaaaaayyyyyy too low… When backfilling tables… it takes days to do so.
The logic for setting the hour credit limit doesn’t make sense. Let’s follow the logic…
If we get 15 credits per second… and only allowed 5,000 credits is an hour… that means we can only run a process for roughly 5.5 mins… right? Since we are tightly restricted with on-demand credits, why impose an hourly credit limit? Sounds to me this API is built on minimal hardware to support the requests.
Thoughts on removing the hourly credit limit?
Thanks,
Mathieu