Private beta · Kubernetes on GKE

Every Kubernetes cluster.
One clear view.

Fleetscope connects your clusters with a read-only agent and shows workloads, usage against requests, logs and problems in one place. No kubectl context switching, no dashboards to build.

Join the private beta See how it works

Install with one Helm command · Read-only · Outbound connection only

prod-sg-1 prod ● Connected

CPU 18.4 cores used

11.6 cores requested but idle

MEMORY 96 GiB used

Healthy · limits within capacity

CPU over timenow 18.4 · peak 26.1 · avg 17.2

1Deploy api-gateway — requests +1.6 cores

Sample data

Running many clusters should not mean many tabs, many contexts and guesswork.

Scattered visibility

Each cluster has its own context, console and dashboards. Finding what is broken across them takes too long.

Requests nobody uses

Pods reserve CPU and memory they never touch. Without usage next to requests, nodes keep growing.

Slow troubleshooting

Events expire after an hour and logs need kubectl access. By the time someone looks, the evidence is gone.

FEATURES

What you see in Fleetscope

Fleet and cluster overview

Every cluster with its health, nodes, pods and agent status. Open one for CPU, memory, pods and nodes at a glance.

Usage against requests

Charts for 5 minutes to 7 days with requests, limits and allocatable, idle requests hatched and overuse in red.

Workload right-sizing

Per-pod CPU and memory, a suggested request from 7 days of usage and a values.yaml patch to copy.

Logs on demand

Read or follow container logs from the browser. Logs stream through the agent and are never stored.

Events and findings

Repeated warnings kept for 7 days, and checks for crash loops, OOM kills, pending pods and node pressure.

Audit log

Who signed in, who changed a cluster and who opened which YAML or log, kept for 90 days.

HOW IT WORKS

Connected in minutes

1
Add a cluster
Name it and pick its environment. Fleetscope gives you a one-time install command.
2
Install the agent with Helm
The agent watches resources read-only and connects outbound. No ports open in your cluster.
3
See your fleet
Workloads appear within seconds; usage charts fill in as metrics arrive every 30 seconds.

# one command per cluster
helm upgrade --install fleetscope-agent \
  oci://ghcr.io/fleetscope-store/charts/fleetscope-agent \
  --namespace fleetscope-system --create-namespace \
  --set controlPlane.url=https://app.fleetscope.store \
  --set clusterName=prod-sg-1 \
  --set bootstrapToken=fs_bt_••••••••••••

SECURITY

Read-only by design

You can review every permission before you install. The agent cannot change anything in your cluster.

RBAC limited to get, list, watch; no exec, no write.
Never reads Secrets or ConfigMaps.
Environment values and password-like arguments are redacted in the agent, before anything leaves the cluster.
Outbound connection only; nothing listens inside your cluster.
Logs stream on request and are not stored.
Every organisation's data is isolated in the database, and access is recorded in the audit log.

ROADMAP

Where Fleetscope is going

Seeing resources is only the start. We are building what single-cluster tools cannot: what changed, what it costs, what is wasted and what will break on upgrade, across every cluster.

Available in the beta

Now

Ready for your clusters today.

Fleet and cluster overview
CPU and memory against requests, limits and allocatable
Workload right-sizing with a values.yaml patch
Logs on demand, never stored
Events, findings and audit log

Coming next

In development now.

Change timeline: deploys, scaling and config changes on every chart
Compare two versions of a workload's spec
Upgrade risk: end-of-support dates and deprecated APIs in use
More checks: missing requests and limits, single replicas, no PDB, failing CronJobs
Cost and waste in money, per cluster, namespace and workload
Alerts to Slack, Telegram, Google Chat and email

Later

Shaped by beta feedback.

Cost per team
Monthly report for engineering leaders
Real GKE billing and Cloud Logging history (optional GCP link)
Teams, roles and single sign-on
Live tail across all pods of a workload

Join the private beta

We are opening Fleetscope to a small number of teams running Kubernetes on GKE. Free during the beta; your feedback shapes what we build next.

Questions? [email protected]

Every Kubernetes cluster.One clear view.