As our traffic ramped up and we approached ~$60/hour in spend, we determined it was necessary to increase both max workers and the hourly spend limit to avoid disruption.
- We submitted a request on the Serverless page (https://console.runpod.io/serverless) to increase both max workers and hourly spend limit. Max workers was increased. Hourly spend limit remained unchanged.
- We opened a support ticket to follow up.
- Several days later, we were asked to file a form (https://share.hsforms.com/2KINjQJPmTmKfgy1p293LVQecykq), which turned out to be effectively identical to the original request in step 1. We completed it immediately.
- After more waiting, we received an email from the Runpod team requesting to set up a “quick call.” We replied asking whether the hourly spend limit could be increased first, as we were already approaching the $80/hour hard cap. The response we received was an out-of-office auto-reply.
- At that point, our spend had reached $72.9/hour, leaving us at immediate risk of hitting the $80/hour limit and triggering a hard stop. With no other option, we emailed again (urgently) to request a temporary increase to prevent service interruption.
This process put us extremely close to an avoidable production outage, despite us acting responsibly and proactively. Frankly, the experience was stressful and left us with a very poor impression at a critical moment.