loki（loki 搜索）

Loki Log Querying

Setup

Loki does not have an HTTPRoute on the internal ingress gateway, so it requires port-forward access (unlike Prometheus and Grafana which are available via DNS).

KUBECONFIG=~/.kube/<cluster>.yaml kubectl port-forward -n monitoring svc/loki-headless 3100:3100 &
export LOKI_URL=http://localhost:3100

Clusters: dev, integration, live

Why port-forward? Loki's gateway component is disabled (replicas: 0) and no HTTPRoute exists for it on the internal ingress gateway. For other services that do have HTTPRoutes (Prometheus, Grafana, Alertmanager), prefer using the internal DNS URLs instead -- see the k8s skill for the full URL table.

Quick Queries

Use the bundled script at .claude/skills/loki/scripts/logql.sh:

# Set URL (Loki requires port-forward -- no internal gateway route)
export LOKI_URL=http://localhost:3100

# Health check
./scripts/logql.sh health

# Search logs by namespace
./scripts/logql.sh tail '{namespace="monitoring"}' --since 15m

# Instant query (metric-style aggregation)
./scripts/logql.sh query 'rate({namespace="monitoring"}[5m])'

# Range query with time window
./scripts/logql.sh range '{namespace="flux-system"}' --start 1h --step 1m --limit 50

# Discover available labels
./scripts/logql.sh labels
./scripts/logql.sh labels namespace

# Find series matching a selector
./scripts/logql.sh series '{namespace="monitoring"}'

Common Operations

Search for Errors

# Errors in a namespace
./scripts/logql.sh tail '{namespace="database"} |= "error"' --since 30m

# Case-insensitive error search
./scripts/logql.sh tail '{namespace="monitoring"} |~ "(?i)error|fail|panic"' --since 15m

# Specific container logs
./scripts/logql.sh tail '{namespace="monitoring", container="prometheus"}' --since 10m --limit 50

Kubernetes Events

# Warning events (via Alloy kubernetes_events)
./scripts/logql.sh tail '{job="integrations/kubernetes/eventhandler"} |= "Warning"' --since 1h

# Events for a specific namespace
./scripts/logql.sh tail '{job="integrations/kubernetes/eventhandler", namespace="database"}' --since 30m

Log Rate Metrics

# Error rate per namespace
./scripts/logql.sh query 'sum by(namespace) (rate({namespace=~".+"} |= "error" [5m]))'

# Log volume by namespace
./scripts/logql.sh query 'sum by(namespace) (bytes_rate({namespace=~".+"}[5m]))'

# Top 5 noisiest pods
./scripts/logql.sh query 'topk(5, sum by(pod) (rate({namespace=~".+"}[5m])))'

Direct curl (alternative)

# Instant query
curl -s "http://localhost:3100/loki/api/v1/query?query=%7Bnamespace%3D%22monitoring%22%7D&limit=10" | jq '.data.result'

# Labels
curl -s "http://localhost:3100/loki/api/v1/labels" | jq '.data'

Reference

For homelab-specific LogQL queries (Flux, Cilium, CloudNative-PG, etc.), see references/queries.md.

Loki Details