Instana

Instana

The community for performance and observability professionals to learn, to share ideas, and to connect with others.

 View Only

Monitoring Argo CD with Instana

By IRFAD K P posted 08/21/26 01:42 AM

  

Introduction

Modern software delivery relies on continuous, automated deployments using GitOps workflows. Argo CD is a widely adopted tool for this purpose, providing declarative, Git-based application delivery to Kubernetes.

Faster deployments also increase the risk of undetected failures. A misconfigured manifest, a broken image tag, or a regression in a canary release can cause production incidents. Without an observability layer, you might not detect a problem until users report it.

IBM Instana addresses this gap. You can use the Instana Argo CD sensor to obtain full visibility into your Argo CD-managed deployments from sync status to progressive delivery strategies, such as canary and blue-green rollouts. Instana correlates deployment events with application performance data. Therefore, you can monitor the full impact of every release.

You can configure the Instana Argo CD sensor, interpret deployment observability data, and integrate performance monitoring with your GitOps workflows.

Why Monitoring Argo CD Matters

Argo CD ensures that your cluster state matches your Git repository. However, it does not provide visibility into whether the deployed application is functioning correctly.

Argo CD alone cannot answer the following questions:

  • Health status of a newly synced version
  • Performance regressions introduced by a canary rollout
  • The commit responsible for a latency spike in production
  • Behavioural consistency of blue-green environments before promotion

IBM Instana addresses these gaps by correlating Argo CD deployment events with real-time metrics, traces, and logs. This capability provides you full visibility into deployment health and application performance across your delivery pipeline.

How the Instana Argo CD sensor works

The Instana Argo CD sensor ships with the Instana host agent and provides continuous monitoring of your Argo CD-managed applications. After you install the agent on your Kubernetes clusters and configure the sensor, the sensor continuously polls the Argo CD REST API to collect synchronization data, application health, and rollout information.

The sensor provides the following capabilities:

  • Real-time sync status for all Argo CD-managed applications
  • A timeline of Argo CD sync events correlated with runtime metrics
  • Full monitoring of Argo Rollouts, including canary and blue-green strategies
  • Application performance monitoring (APM) correlation of error rates, latency, logs, and traces for each deployment stage

Step 1: Install the Instana agent on Kubernetes

Before you configure the Argo CD sensor, install the Instana host agent in your Kubernetes cluster. The agent includes the Argo CD sensor; you do not need to install the sensor separately.

Follow the Instana agent setup guide for Kubernetes to install the agent.

Note: Deploy the agent in the same cluster that Argo CD manages. Instana supports Argo CD installations within a single cluster that may manage one or more target Kubernetes clusters.

Step 2: Set Up Argo CD Access Credentials

The Instana Argo CD sensor reads data through the Argo CD REST API and requires credentials with read-only access. Create a dedicated Argo CD user and restrict its access to only the applications and clusters you want to monitor. This follows the principle of least privilege and limits the sensor’s access to only the data it needs.

Use the following Argo CD documentation to configure access credentials:

Step 3: Configure the Argo CD Sensor

To enable monitoring, add the Argo CD sensor configuration to the configuration file for the Instana agent. After you start the agent and prepare the credentials, open the configuration file at <agent-install-dir>/etc/instana/configuration.yaml and add the following sensor block:

# Add actual Argo CD sensor configuration block here

Note: Replace <agent-install-dir> with the path to your Instana agent installation directory.

# Argo CD Sensor
com.instana.plugin.argocd:
enabled: true
poll_rate: 5 # Polling interval in seconds (default: 5)
url: 'https://argocd-server.argocd'
username: 'instana-readonly'
password: 'your-password'
clusters:
- 'demo-cluster1'
- 'production-cluster'

Configuration reference:

Step 4: View Argo CD Sync Data in Instana

After you configure the Argo CD sensor, you can view sync status and application health directly in Instana.

In Instana UI, go to Platforms > Kubernetes and select the Clusters tab. For each Argo CD-managed cluster, Instana displays a sync status indicator that shows the number of applications that are out of sync.

Note: If the sync status indicator does not appear, verify that the sensor configuration is correct and that the Instana agent is running in the same cluster that Argo CD manages.

The Kubernetes cluster list showing Argo CD sync status at a glance.

Select a cluster to open the Cluster Dashboard. The dashboard displays a timeline of Argo CD sync events alongside Kubernetes performance metrics, which helps you correlate deployment events with cluster performance.

Select the Argo CD sync status indicator to see a detailed list of all Argo CD-managed applications for that cluster. The application list provides visibility into every deployment and includes the following information:

  • Sync status: Whether the application is Synced or Out of Sync with Git
  • Namespace: The namespace where the application is deployed
  • Git repository and branch: The source repository and branch for the application
  • Last sync: The time of the last successful sync

From this view, you can select Launch Argo CD Interface to open the Argo CD web UI, or select any branch name to navigate directly to the source Git repository.

The cluster dashboard correlates Argo CD sync events with runtime metrics on the same timeline.

Step 5: Monitor Argo Rollouts

If you use Argo Rollouts for progressive delivery, Instana also monitors your rollouts and provides visibility into canary and blue-green deployments in real time.

From the Cluster Dashboard, select the Argo Rollouts tab. Each rollout card shows the following information:

  • Rollout name, namespace, and deployment strategy (canary or blue-green)
  • Health status: Healthy, Progressing, Suspended, or Degraded
  • Replica counts: ready versus desired
  • A summary bar at the top of the tab displays a health summary of all rollouts in the cluster.

Note: If you install the Instana APM agent, each rollout card also displays error rate and latency for the rollout.

The Argo Rollouts tab showing all active rollouts with health status, strategy, and APM metrics.

Canary Rollout Visibility

Canary deployments shift traffic gradually to a new version so you can monitor for issues before full promotion. Instana provides visibility into every step of this process.

Select a canary rollout card to see the following information:

  • Step progress: Current step in the rollout plan, for example, Step 3 of 7
  • Traffic weight distribution: Percentage of traffic routed to the canary versus the stable variant
  • Side-by-side APM comparison: Error rates and latency for the canary versus the stable variant

You can use this data to decide whether to proceed with the rollout or roll it back.

The canary rollout detail view — traffic weight, step progress, and a side-by-side performance comparison between canary and stable versions.

Blue-Green Rollout Visibility

Blue-green deployments run two identical environments — active and preview — simultaneously. When you are confident that the preview version is healthy, you promote it to receive production traffic.

Select a blue-green rollout card to see the following information:

  • Active versus preview indicators: Which version serves live traffic and which is staged
  • Replica status: Replica counts for both the active and preview environments
  • Performance metrics: Error rates and latency for each version

If you install the Instana APM agent, you can also view correlated logs and traces:

Use this view to compare the actual performance of the preview version against the active version before you shift traffic permanently.

Blue-green rollout detail — clearly identifying active and preview versions with replica and performance data.

APM correlation in a blue-green rollout — compare services, access logs, and analyse traces to validate readiness before promotion.

Automated rollback with the Instana metric provider plugin

You can configure the Instana metric provider plugin to trigger automated rollback when Instana detects a performance degradation in a canary release.

Instana provides a plugin for Argo Rollouts that enables automated rollback based on live Instana metrics. If your application performance monitoring (APM) data shows that a canary release degrades performance below defined thresholds, the plugin automatically aborts the rollout without requiring manual intervention.

The rollouts-plugin-metric-instana plugin allows you to define Instana metrics as analysis criteria. If the canary fails the analysis, Argo Rollouts rolls back automatically.

To configure automated rollback, see the rollouts-plugin-metric-instana plugin documentation.

Instana and Argo CD capabilities

The following table compares capabilities with and without the Instana sensor:

Conclusion

Argo CD and IBM Instana serve complementary functions in a Kubernetes delivery workflow. Argo CD ensures your Kubernetes cluster reflects your Git state. Instana ensures you know whether that state is healthy and performing within expected thresholds.

By combining GitOps automation with full-stack observability, you can deploy more frequently and with greater reliability — whether you run a standard sync, a canary rollout with traffic shaping, or a blue-green promotion in production.

If you use Argo CD on Kubernetes, adding the Instana sensor requires a small configuration change and provides immediate visibility into deployment health.


#Agent
#BusinessObservability
#CustomDashboards
#EUM
#Infrastructure
#IntelligentRemediation
#Kubernetes
#Cloud
#Remediation

0 comments
22 views

Permalink