atscalectl status

The status command shows the current state of an AtScale deployment running on Kubernetes, including pod readiness, restart counts, pod age, AtScale version, resource usage, and recent error messages from unhealthy pods. This is the first command to run when investigating an issue. It is safe to execute at any time because it only reads information from the Kubernetes cluster.

Prerequisites

status requires the following Kubernetes permissions:

• pods get/list
• pods/log get
• secrets list. Required only when the --release-name flag isn't specified. This is required to detect the Helm release.
• (Optional) metrics.k8s.io/pods get. This is required only for resource usage information.

Usage

atscalectl status [flags]

Flags

status supports all global flags. The following may be especially useful:

• --context
• --namespace
• --release-name
• --output

In addition to the global flags, status supports the following command-specific flags.

Flag	Default	Description
--no-resources	false	Skips CPU and memory information.
--no-logs	false	Skips unhealthy pod log analysis.

Output

By default, status outputs a table that contains up to three sections:

• Health
• Resources
• Unhealthy pod logs

Note

When using structured output (atscalectl status -o json), the JSON data is still written to standard output, while diagnostic messages are written to standard error.

The table sections and their contents are described in detail below.

Health

The health section contains one row per Kubernetes pod. Each row contains the following columns.

Column	Description
COMPONENT	The AtScale component associated with the pod; for example, engine, keycloak, database. Note that database is listed only when PostgreSQL is deployed inside Kubernetes; external PostgreSQL deployments do not show a database pod.
POD	The Kubernetes pod name.
STATUS	The current pod/container status. Container failure states such as CrashLoopBackOff are shown when available.
READY	The number of ready containers compared to total containers.
RESTARTS	Total container restarts.
AGE	Time since pod creation.
VERSION	The AtScale version extracted from the container image.
DEPLOYED	Pod creation time in UTC.

Resources

Resource information includes:

• CPU requests
• CPU limits
• Memory requests
• Memory limits
• Current CPU and memory usage

Note

Current usage is provided by Kubernetes metrics-server. If metrics-server is unavailable, resource requests and limits are still displayed, but current usage values show -.

Unhealthy pod logs

When a pod is not fully ready, status displays a small set of matching log lines from that pod. The output is filtered for common error indicators, such as:

• error
• critical
• warn
• fatal
• panic
• exception

Note

This section is intended as a quick diagnostic hint. For complete troubleshooting data, run atscalectl logs collect. For more information, see atscalectl logs collect.

Exit codes

status returns the following exit codes.

Code	Description
0	All pods are healthy.
1	At least one pod is not fully ready, or the command failed.

Usage examples

Check the default deployment

atscalectl status

Check another namespace

atscalectl status --namespace <namespace>

Use JSON output for automation

atscalectl status -o json

Export status information

atscalectl status -o csv > status.csv

Skip metrics collection

atscalectl status --no-resources

Health check

if atscalectl status --context <context> > /dev/null; then
  echo "healthy"
else
  echo "not healthy"
fi

Troubleshooting

All usage columns show -

This may be caused by metrics-server not being installed or available. However, resource requests and limits should still be available. If you do not require metrics data, run:

atscalectl status --no-resources

No pods found

If you receive the following error, then AtScale is installed in another namespace, or the Helm release name cannot be detected.

no pods found in namespace "atscale"

Try running the following to determine the AtScale namespace:

atscalectl status --namespace <namespace>

Or run the following to determine the release name:

atscalectl status --release-name <release>

Multiple Helm releases found

If you receive the following error, it means there are multiple Helm releases in the same namespace.

could not auto-detect release name: multiple helm releases found in namespace "atscale"

If this occurs, you must specify the release explicitly:

atscalectl status --release-name <release>