Serverless scaling
I'm considering using runpod for commercial use. I need reliable, relatively cheap scaling for this to work, but ive heard that at least a few months ago serverless was very unreliable, i.e. not allocating GPUs for hours or days at a time. I don't want to have to figure out how to deploy on runpod just to realize that it's unreliable. What is your take on this right now? Any evidence that these problems have been fixed?
2 Replies
Why did you ask here and also in #🎤｜general ? You should only post in 1 place and not multiple places.
you can run active workers for your base workload and use flex workers to handle spikes, this depends on what kind of throughput you have