Service deploys fine, works for a few hours days, then we get: "Application Failed To Respond".

We're a bit stumped here... Throughout September we've had some issues where our API apparently goes down, but no logs are provided, everything 'looks' fine (no errors, warnings etc). Then, we'll get messages from our users complaining about timeouts, application not responding, and when we visit our fast api docs, it is indeed down. Once we restart, or redeploy, we're good again for a day or so until it happens again. We've turned off the legacy proxy on the API service but it hasn't helped much.