-
Notifications
You must be signed in to change notification settings - Fork 0
GKE Mesh
Module flag: gke_mesh
Enabled by default: yes
The GKE Mesh module surfaces service-to-service traffic topology from Istio or Anthos Service Mesh (ASM) telemetry. It infers edges between services based on actual observed traffic rather than configuration intent.
GCP_PROJECT_ID=my-project ./aura-tracker-gcp --modules gke_meshgcloud services enable container.googleapis.com monitoring.googleapis.com| Role | Tools that need it |
|---|---|
roles/container.viewer |
Cluster and workload context |
roles/monitoring.viewer |
Istio telemetry metrics |
Returns service-to-service traffic edges inferred from Istio/ASM telemetry metrics.
Parameters:
| Name | Required | Description |
|---|---|---|
project_id |
yes | GCP project ID |
location |
yes | Region or zone of the GKE cluster |
cluster_name |
yes | Cluster name |
namespace |
no | Filter to a specific namespace |
hours |
no | Lookback window in hours (default: 1) |
Returns: Directed edges between services (source → destination), request rates, error rates, P50/P99 latency. Services with no traffic in the window are omitted.
Note: Requires Istio or ASM to be installed and emitting telemetry to Cloud Monitoring. If mesh telemetry is not available, the tool returns an empty edge set.
Traffic map
"Show me the service-to-service traffic topology for cluster
prod-clusterinus-central1over the last hour."
Error path tracing
"Which services in the
checkoutnamespace ofprod-clusterare receiving error responses from their upstream dependencies?"
Latency hotspot
"Which service-to-service calls in
prod-clusterhave P99 latency above 500ms?"
| Module flag | Why you'd combine it |
|---|---|
gke_workloads |
Match traffic edges to Kubernetes workload specs |
archgraph |
Include mesh topology in the full project architecture graph |
monitoring |
Fetch raw Istio metrics by metric descriptor |