Specify minimum replica of 1 for prediction service in Twilio labs

RedHatGov / serverless-workshop-dashboard

Lab guides for a workshop on OpenShift Serverless (Knative)

Apache License 2.0

5 stars 4 forks source link

Specify minimum replica of 1 for prediction service in Twilio labs #30

Closed theckang closed 3 years ago

theckang commented 3 years ago

Twilio has a hard timeout of 15 seconds. When no pod is running, it seems it can take > 15 seconds for knative to spin the pod up and make it ready to respond to requests.