Spinnaker Integration
Deployment pipeline metrics, canary analysis results, and rollback event monitoring. Correlate every Spinnaker deployment with production metrics and automate rollback decisions.
How It Works
Enable Spinnaker Echo Webhooks
Configure Spinnaker Echo to forward pipeline execution events to TigerOps. Add the TigerOps webhook endpoint to your echo-local.yml. All pipeline start, stage complete, and pipeline complete events are forwarded automatically.
Configure Kayenta Metric Store
Point Spinnaker Kayenta at your TigerOps remote-read endpoint as a metric store. This enables TigerOps metrics to be used directly in automated canary analysis judgments alongside your existing metric sources.
Add TigerOps Stage to Pipelines
Insert the TigerOps Change Event stage into your Spinnaker pipelines. The stage fires a structured deployment event to TigerOps at the point you choose — before bake, after deploy, or as a pipeline notification.
Correlate Deployments with Metrics
TigerOps overlays Spinnaker deployment events on your application dashboards and automatically flags metrics that degraded after each deploy, integrating canary analysis scores into a unified change timeline.
What You Get Out of the Box
Pipeline Execution Metrics
Track pipeline start-to-finish duration, per-stage execution times, and success and failure rates per pipeline and application. TigerOps alerts when pipeline execution times regress beyond historical baselines.
Canary Analysis Score Tracking
Ingest Kayenta canary analysis scores and judgments as time series metrics. TigerOps stores canary pass/fail history per service and highlights when canary thresholds are trending toward failure across multiple deployments.
Rollback Event Monitoring
Capture automatic and manual rollback events with the triggering reason, the rolled-back version, and the target version. TigerOps correlates rollbacks with the metric anomalies that preceded them for post-incident analysis.
Stage Failure Analysis
Monitor which Spinnaker stages fail most frequently — bake, deploy, resize, run job — and surface the error messages and affected clusters. TigerOps groups failures by cloud provider, region, and stage type.
Deployment Frequency & Lead Time
Compute DORA deployment frequency and change lead time from Spinnaker pipeline completion events. Track trends per application and team to measure the impact of pipeline optimizations over time.
Gate & Manual Judgment Timing
Track how long deployments wait at manual judgment gates, approval stages, and traffic management gates. TigerOps surfaces bottlenecks in your pipeline approval workflows that slow overall delivery throughput.
Spinnaker Echo Webhook Configuration
Configure Spinnaker Echo to forward pipeline execution events to TigerOps.
# echo-local.yml — Spinnaker Echo webhook configuration for TigerOps
# Place in /opt/spinnaker/config/ or your Halyard config directory
rest:
enabled: true
endpoints:
- wrap: false
url: https://ingest.atatus.net/api/v1/spinnaker/events
# Use a header to pass the TigerOps API key
headers:
Authorization: "Bearer ${TIGEROPS_API_KEY}"
# Forward all event types for full pipeline visibility
# Supported: PIPELINE, STAGE, TASK, MANUAL_JUDGMENT, ARTIFACT
eventTypes: [] # empty = forward all
# ─── Kayenta metric store for TigerOps (kayenta-local.yml) ────────
#
# metricStores:
# tigerops:
# enabled: true
# endpoint: https://query.atatus.net/api/v1/kayenta
# apiKey: ${TIGEROPS_API_KEY}
#
# canaryAnalysis:
# defaultMetricStore: tigerops
#
# ─── Example pipeline stage JSON ──────────────────────────────────
# {
# "type": "webhook",
# "name": "Notify TigerOps — Deploy Start",
# "url": "https://ingest.atatus.net/api/v1/spinnaker/events",
# "method": "POST",
# "headers": { "Authorization": "Bearer ${TIGEROPS_API_KEY}" },
# "payload": {
# "event": "deploy_start",
# "service": "${parameters.service}",
# "environment": "${parameters.environment}",
# "version": "${trigger.buildInfo.number}"
# }
# }Common Questions
Which Spinnaker versions and cloud providers does TigerOps support?
TigerOps supports Spinnaker 1.28+ via Echo webhooks. It works across all Spinnaker cloud providers — AWS, GCP, Azure, Kubernetes, and more — because event forwarding happens at the Echo layer regardless of the underlying cloud provider.
How does TigerOps integrate with Kayenta canary analysis?
TigerOps exposes a remote-read endpoint compatible with Prometheus as a Kayenta metric store. This means your canary analysis can judge against TigerOps metrics. Additionally, TigerOps ingests Kayenta judgment results as structured events for historical tracking.
Can TigerOps trigger a Spinnaker rollback when anomalies are detected?
Yes. TigerOps can call the Spinnaker Gate API to trigger a rollback pipeline when its AI correlation engine detects a metric anomaly that is causally linked to a recent Spinnaker deployment. This requires configuring a Spinnaker API key in TigerOps.
Does TigerOps support Managed Pipeline Templates (MPT)?
Yes. TigerOps tracks events from pipelines created from MPTs the same way as regular pipelines. The template name and version are included in the event metadata so you can correlate pipeline performance with specific template versions.
How does TigerOps handle parallel pipeline branches in Spinnaker?
TigerOps tracks each parallel branch as a child execution. The pipeline execution tree is reconstructed from Echo events, giving you a complete view of parallel branch success rates and the critical path duration for complex deployment pipelines.
Close the Loop Between Canary Scores and Production Reality
Pipeline metrics, canary analysis integration, and automated rollback correlation. Set up in minutes.