This page contains usage quota and limits that apply when using Knative serving.
Knative serving is subject to the Google Kubernetes Engine quotas and limits.
The number of Knative serving resources is limited by the configuration of the cluster as well as other dependencies. The following limits are recommended limits for a properly scaled Kubernetes Engine cluster.
| Resource | Description | Limit | Can be increased | Scope |
|---|---|---|---|---|
| Services | Maximum number of services | 150 | No | per cluster |
| Revisions | Maximum number of revisions | 300 | No | per cluster |
| Timeout | Maximum time before timeout for 0.16.0-gke.1 and later | 24 hours | No | per request |
| Maximum time before timeout for 0.15.0-gke.3 and earlier | 900 seconds | No | per request |
Other resource limitations are imposed by the configuration of the Kubernetes Engine cluster that the services are running in. For example, you cannot request more memory than is available in the nodes in the cluster.