Linkerd Integration
Ingest Linkerd golden signals with zero application instrumentation. Monitor mTLS coverage, inter-service traffic, and service-level SLOs with AI-powered anomaly detection.
How It Works
Enable Linkerd Viz Extension
Install the Linkerd viz extension in your cluster. It exposes a Prometheus-compatible metrics endpoint that TigerOps scrapes for per-route and per-service golden signals.
Configure TigerOps Scrape
Add the TigerOps scrape config pointing to the Linkerd viz metrics endpoint. All success rates, latency percentiles, and request volumes begin flowing within minutes.
Define SLOs per Service
Set success rate and P99 latency SLOs for each Linkerd meshed service. TigerOps fires alerts when a service breaches its SLO and surfaces the upstream caller.
Correlate with Infrastructure
TigerOps correlates Linkerd service-level degradations with node pressure, pod restarts, or downstream database slow queries for full-stack root cause analysis.
What You Get Out of the Box
Golden Signal Dashboards
Per-service and per-route success rate, request rate, and P99/P999 latency dashboards built from Linkerd proxy metrics. No application code changes required.
mTLS Visibility
Track mTLS coverage across your mesh. TigerOps alerts when a workload drops to plaintext connections, indicating a misconfigured proxy or missing annotation.
Topology & Traffic Maps
Visualize inter-service traffic flow with request volume and error rates on each edge. Identify which upstream callers are impacting a degraded downstream service.
Retries & Timeouts
Monitor Linkerd retry budgets, timeout counts, and circuit breaker trip rates per service route. Excessive retries are surfaced as early warning indicators.
Control Plane Health
Track Linkerd control plane component health — destination, identity, and proxy-injector — with memory, CPU, and reconciliation error metrics.
AI Anomaly Detection
TigerOps AI learns your baseline success rates and request volumes per service. Deviations trigger automated root cause investigations with correlated signals.
Linkerd Viz + TigerOps Scrape Config
Enable Linkerd viz and point TigerOps at the metrics endpoint.
# Step 1: Install Linkerd viz extension
linkerd viz install | kubectl apply -f -
# Step 2: Verify viz metrics are available
linkerd viz dashboard &
# Metrics endpoint: http://prometheus.linkerd-viz.svc.cluster.local:9090
# Step 3: TigerOps scrape config (prometheus-scrape.yaml)
apiVersion: v1
kind: ConfigMap
metadata:
name: tigerops-scrape-config
namespace: tigerops
data:
scrape.yaml: |
scrape_configs:
- job_name: linkerd-viz
honor_labels: true
kubernetes_sd_configs:
- role: pod
namespaces:
names: [linkerd-viz]
relabel_configs:
- source_labels: [__meta_kubernetes_pod_label_component]
action: keep
regex: prometheus
metrics_path: /federate
params:
match[]:
- '{job="linkerd-proxy"}'
- '{job="linkerd-destination"}'
- '{job="linkerd-identity"}'
remote_write:
- url: https://ingest.atatus.net/api/v1/write
bearer_token: "${TIGEROPS_API_KEY}"
write_relabel_configs:
- source_labels: [__name__]
regex: 'response_total|request_total|response_latency.*|tcp_.*'
action: keepCommon Questions
Does TigerOps require changes to Linkerd-meshed pods?
No. TigerOps scrapes metrics from the Linkerd viz extension Prometheus endpoint. No changes to application pods, annotations, or proxy configuration are required.
Which Linkerd versions does TigerOps support?
TigerOps supports Linkerd 2.10 and later, including Linkerd Edge releases. The viz extension must be installed for full metrics access. Linkerd CNI mode is also supported.
Can TigerOps alert on mTLS failures?
Yes. TigerOps monitors the tls_not_provided_by_remote_total counter per service pair. When a previously mTLS-secured route starts receiving plaintext traffic, an alert fires automatically.
How does TigerOps handle Linkerd multicluster setups?
Each cluster runs its own TigerOps scraper. Metrics are tagged with a cluster label so you can compare service health across clusters in unified dashboards and correlated alerts.
Can I use TigerOps alongside Linkerd Viz dashboards?
Yes. TigerOps and Linkerd Viz serve complementary roles. Linkerd Viz provides real-time topology views while TigerOps adds historical trending, AI anomaly detection, and cross-stack correlation.
Full Service Mesh Visibility with Zero App Changes
Golden signals, mTLS monitoring, and AI anomaly detection for Linkerd. Ready in minutes.