Ruptura runs in your cluster on top of the Prometheus you already have. It turns telemetry into clear conclusions: the failure, the time left, the evidence, and the next step.
GET /api/v2/failures { "target": "demo-shop/Deployment/step-leak", "failureMode": "saturation", "kind": "predicted", "state": "degraded", "etaSeconds": 626, "recommendedAction": "Review the memory limit or restart the pod before it is killed", "evidence": ["Memory at 75% of its limit"] }
One structured result per failure, ready for your alerts and your tools.
A rule like "memory above 90%" says a pod is in trouble, not whether it will be killed in two minutes or never. Ruptura answers the question that matters.
One result per failure, with the target, the failure mode, the time left and the evidence. No new dashboard to watch.
Every workload is compared with its own history and its siblings, not with a global threshold.
Predicted and detected failures are kept apart, and Ruptura reports the data it cannot see.
Choose the level to watch, how deep to collect, and where results go.
One custom resource, one operator, no external database. On OpenShift, results reach the cluster Alertmanager.
Apache 2.0. Runs entirely in your cluster. Your telemetry never leaves it.
Ruptura sits on top of your existing monitoring stack. Nothing to replace.
Reads Prometheus, OTLP and the Kubernetes API.
Maps every metric to the workload it describes.
Forecasts each workload and spots anomalies against its own history.
Names the failure and sends it to Alertmanager, dashboards and your tools.
Predicted failures carry the time left. Detected failures report what is happening now.
Full rules in the failure modes reference.
Pick the path that fits your environment.
oc apply -f https://raw.githubusercontent.com/\ benfradjselim/ruptura/main/install/operator.yaml
Then create a RupturaInstance. On Kubernetes, set your Prometheus URL in it.
helm install ruptura \ oci://ghcr.io/benfradjselim/charts/ruptura \ -n ruptura-system --create-namespace \ --set thanos.url=<your-prometheus>
Engine only. See the README for all options.
git clone https://github.com/\ benfradjselim/ruptura cd ruptura docker compose up --build
Open localhost:8081 and try it on demo workloads.
Ruptura is in active development and not yet recommended for production. Try it on a test cluster and help shape the roadmap.