3106adec39
Previously, we used 307 Internal Redirect to tell k8s that we're still waiting for the healthcheck window to fill up. Unfortunately, k8s assumes 3XX as a success, which wrongly tells it that the service is ready. Changing to 503 causes the expected probe failure to appear when the healthcheck state is not yet determined. |
||
|---|---|---|
| .. | ||
| client | ||
| governance | ||
| program | ||
| sdk | ||
| Cargo.lock | ||
| Cargo.toml | ||
| rust-toolchain | ||