Environments
An Environment allows you to group a set of services having a specific version, usually based on a branch of your repository. For example, you can have one Production environment (all services pointing to the main branch), one Staging environment (all services pointing to the staging branch), and so on. You run your application on an environment.
Environment List
Navigate to Environments from the left menu to view all environments in your organization.
| Column | Description |
|---|---|
| Env Name | The name assigned to your environment. Must be unique and cannot contain special characters. |
| Last Updated | When the environment was last created or updated. |
| Cluster | The Kubernetes cluster on which the environment is deployed. |
| Project | The project (application blueprint) deployed on the environment. |
| Created By | The user who deployed the environment. |
| Status | Current phase of the environment (see below). |
Environment Statuses
| Status | Description |
|---|---|
| Initializing | One or more services are being deployed — pipelines are running for application services or datasources are being provisioned. |
| Active | All pipelines have completed and all services are healthy. |
| Updating | Services are being added, re-deployed, auto-triggered (continuous CI/CD), or deleted from the environment. |
| Error | A pipeline has failed for an application service or a datasource failed to deploy. |
| Destroying | The environment and its resources are being destroyed. |
| Destroyed | The environment and all its resources have been fully destroyed. |
| Decommissioning Resources | Environment resources are being scaled down (stop operation in progress). |
| Unavailable | All services in the environment have been stopped. Pods are scaled to zero. |
| Provisioning Resources | Environment resources are being scaled back up (start operation in progress). |
| Destroy Failed | The destroy job for the environment has failed and needs to be re-triggered. |
| Service Unavailable | All services have been deleted from the environment. |
| Unknown | The environment state cannot be determined — services are in transitional or unresolvable statuses. |
| Partially Active | Some services are running while others remain stopped after a start operation. |
| Partially Stopped | Some services have been stopped while others remain running after a stop operation. |
Service Health Statuses
Each service within an environment has its own health status:
| Status | Description |
|---|---|
| Healthy | The service is running and passing health checks. |
| Running | The service pods are running. |
| Progressing | The service is being deployed or updated. |
| Starting | The service is being scaled up. |
| Stopping | The service is being scaled down. |
| Stopped | The service pods have been scaled to zero. |
| Degraded | The service is partially available or experiencing issues. |
| Missing | The expected service resources are not found on the cluster. |
| Release Failed | The Helm release or deployment for the service failed. |
| N/A | Health status has not yet been determined. |
Environment Overview
Clicking on an environment opens the Environment Overview, which shows a summary of your deployed setup — including all application services, datasources, and their current statuses.
Labels & Guardrails
You can view custom labels and any associated guardrail policies within the environment overview section. Labels are key-value pairs used for categorization, and guardrails are automated policies that can trigger actions like scaling.
Environment Actions
From the environment list or overview, you can perform the following actions via the three-dot menu:
| Action | Description |
|---|---|
| Clone | Duplicate the environment with all its configurations. See Cloning. |
| Start / Stop | Scale all services up or down. See Start & Stop Environment. |
| Edit | Modify environment settings or service configurations. See Editing or Destroying. |
| Destroy | Permanently remove the environment and all its resources. See Editing or Destroying. |
| Reconcile | Sync the environment state in Atmosly with the actual Kubernetes cluster state. Useful when the displayed status does not match reality. |
Use Reconcile if an environment's status appears stuck or incorrect. It fetches the live state from the cluster and updates all service statuses accordingly.
Health Issues & AI Debugging
Atmosly continuously monitors your environments for health issues and surfaces them in real time. Issues are detected automatically and reported per environment and namespace.
Accessing Health Issues
You can access health issues from multiple places:
- Top navigation bar — A badge displays the number of active issues for the current environment. Click it to open the Health Issues panel.
- Environment cards — Each environment card on the dashboard shows the issue count. Click to view details.
- Services Overview — Open the Health Issues panel from within any environment's service view.
Issue Details
Each detected issue includes:
| Field | Description |
|---|---|
| Title | A short summary of the issue (e.g., "Pod in CrashLoopBackOff"). |
| Description | Detailed explanation of what was detected. |
| Category | The issue category — performance, memory, network, storage, security, reliability, deployment, resource, or configuration. |
| Resource | The affected Kubernetes resource (Pod, Service, Deployment, ConfigMap, Secret, etc.). |
| Namespace | The namespace where the issue was detected. |
| Severity | The severity level — Critical, High, Medium, Warning, Low, or Info. |
| Status | Current state — Active, Investigating, Acknowledged, Pending, or Resolved. |
| Detected At | When the issue was first detected. |
Filtering & Tabs
The Health Issues panel provides two tabs:
- Active Issues — All currently unresolved issues.
- Resolved Issues — Issues that have been automatically or manually resolved.
You can filter issues by severity and category to focus on what matters most.
AI-Powered Root Cause Analysis
For any active issue, you can trigger an AI-Enhanced Root Cause Analysis to get an in-depth diagnosis. Click the Root Cause Analysis button on an issue to generate a report that includes:
| Section | Description |
|---|---|
| Root Cause | The most likely cause of the issue. |
| Contributing Factors | Additional factors that may be involved. |
| Impact Assessment | How the issue affects your environment and services. |
| Recommended Actions | Step-by-step remediation suggestions. |
| Timeline of Events | A chronological view of events leading to the issue. |
| Diagnostic Information | Relevant logs, metrics, and resource states. |
| Confidence Score | How confident the analysis is in the identified root cause. |
| Estimated Resolution Time | Approximate time to resolve the issue. |
Use AI Root Cause Analysis for complex issues like intermittent pod crashes or network failures — it correlates multiple signals across your cluster to identify the underlying problem.
Notifications
Health issues can trigger notifications via email and Slack when critical or high-severity issues are detected. Configure which health issue notifications you want to receive from the Notification Settings page.
Auto-Resolution
Issues are automatically marked as resolved when the underlying problem is no longer detected in the cluster. Resolved issues remain visible in the Resolved tab for reference.
Permissions
To manage environments, your role must have the appropriate Environment permissions:
| Permission | Description |
|---|---|
| View environments | View the environment list and details |
| Add environment | Create new environments |
| Change environment | Update, start, stop, or reconcile environments |
| Delete environment | Destroy environments |