All Integrations
ContainersHelm 3 + Kubernetes API

Helm Integration

Track every Helm release event, correlate chart deployments with metric changes, and get instant AI root cause analysis when an upgrade causes a service regression.

Setup

How It Works

01

Install the TigerOps Agent Chart

Add the TigerOps Helm repository and install the agent chart into your cluster. It registers a post-install/post-upgrade hook that ships release events automatically.

02

Annotate Deploy Events

Each helm install, upgrade, rollback, and uninstall is captured as a structured event in TigerOps — including chart version, values hash, namespace, and initiating user.

03

Correlate with Metric Changes

TigerOps overlays release events on your dashboards. When a deploy coincides with a latency spike or error rate increase, the AI engine flags the correlation immediately.

04

Alert on Failing Releases

Detect stuck upgrades, failed hooks, and pending rollbacks. TigerOps fires an alert if a release stays in a non-deployed state beyond your configured timeout.

Capabilities

What You Get Out of the Box

Release Lifecycle Events

Every helm install, upgrade, rollback, and uninstall is captured as a timestamped event with full metadata — chart name, version, namespace, and values diff.

Deploy Correlation

TigerOps automatically overlays Helm release markers on all relevant metric dashboards so you can visually tie a deployment to any change in service behavior.

Rollback Detection

When a helm rollback is executed, TigerOps creates an incident annotation and notifies your alerting channels with the triggering release and the reverted version.

Chart Version Drift

Track which chart versions are deployed across all namespaces and clusters. Get alerts when a workload falls behind the canonical chart version in your repository.

Hook Failure Tracking

Pre-install, pre-upgrade, post-install, and post-upgrade hook failures are surfaced as incidents with the hook job logs attached for immediate diagnosis.

AI Regression Analysis

When metrics degrade after a release, TigerOps AI compares the values diff, changed container images, and resource limit changes to identify the likely cause.

Configuration

Install the TigerOps Agent via Helm

One command installs the agent and begins tracking all Helm releases in your cluster.

helm-install.sh
# Add the TigerOps Helm repository
helm repo add tigerops https://charts.atatus.net
helm repo update

# Install the TigerOps agent with Helm release tracking enabled
helm install tigerops tigerops/agent \
  --namespace tigerops \
  --create-namespace \
  --set apiKey="${TIGEROPS_API_KEY}" \
  --set helmWatcher.enabled=true \
  --set helmWatcher.namespaces="{production,staging}" \
  --set helmWatcher.alertOnStuckUpgrade=true \
  --set helmWatcher.stuckUpgradeTimeoutMinutes=15

# Verify the agent is watching releases
helm list -n tigerops
kubectl get pods -n tigerops

# The agent values file (values.yaml)
# helmWatcher:
#   enabled: true
#   namespaces:
#     - production
#     - staging
#   excludeReleases:
#     - "test-*"
#     - "preview-*"
#   alertOnStuckUpgrade: true
#   stuckUpgradeTimeoutMinutes: 15
#   correlationWindow: 10m   # look-back window for metric correlation
#
# remoteWrite:
#   endpoint: https://ingest.atatus.net/api/v1/write
#   bearerToken: "${TIGEROPS_API_KEY}"
#
# events:
#   endpoint: https://ingest.atatus.net/api/v1/events
#   clusterName: prod-us-east-1
FAQ

Common Questions

Does TigerOps require changes to existing Helm charts?

No. The TigerOps agent installs as a separate chart and uses the Kubernetes API to watch Helm secret resources (which Helm uses to store release state). No modifications to your existing charts are required.

Which Helm versions are supported?

TigerOps supports Helm 3.x. Helm 2 (Tiller-based) is not supported. The agent reads Helm 3 release secrets from the kube-system or target namespace as configured.

Can I filter which releases are tracked?

Yes. You can include or exclude releases by namespace, release name prefix, or label selector in the TigerOps agent Helm values. This lets you focus on production namespaces and ignore development environments.

How does TigerOps handle multi-cluster Helm deployments?

The TigerOps agent is installed per cluster. Each agent tags its events with a configurable cluster identifier, so release events from multiple clusters are unified in the same TigerOps workspace with full cluster context.

Can TigerOps trigger alerts if a Helm upgrade is stuck?

Yes. You can configure a maximum upgrade duration threshold. If a release stays in the pending-upgrade or pending-install state beyond that threshold, TigerOps fires an alert and includes the current hook job status.

Get Started

Know Instantly When a Helm Release Breaks Something

Deploy event correlation, rollback detection, and AI regression analysis. Install in 2 minutes.