Platform · Monitor

Setting up monitoring, means zero effort.

Infrastructure and your code, covered from day one. No agents to install. No dashboards to build from scratch.

Built-in

Open-source observability, wired up for you.

The full Grafana stack — Loki for logs, Prometheus for metrics, Grafana for dashboards and alerting — running inside your environment.

Logs
Loki
Stores every line your services and infrastructure write — searchable from Grafana on day one.
Metrics
Prometheus
Per-service request rate, latency, error rate and resource usage — scraped and stored automatically.
Dashboards + alerts
Grafana
Pre-built dashboards. Alerts to Slack, MS Teams, Email, PagerDuty and any webhook.

Alert toSlackMS TeamsEmailPagerDutyWebhook

Coverage

Full-stack coverage, infrastructure and your code.

Infrastructure

Cluster, nodes, networking and databases

Logs and metrics enabled by default. Stored in your cloud's managed observability store (CloudWatch / Cloud Monitoring / Azure Monitor) and wired up as Grafana data sources.

Cloud-managed log + metric stores attached
Grafana data sources pre-configured
Alerts routed through Grafana

Your services

Logs and metrics, automatically

Write logs to stdout / stderr — Loki picks them up. Expose Prometheus metrics on /metrics — they get scraped automatically.

stdout / stderr → Loki, visible in Grafana
/metrics → Prometheus, default RED + USE charts
Crash + restart history, per service

Custom metrics

Track any metric, alert on any of them.

TCP connections, notifications sent, emails delivered, queue depth — expose anything you want on /metrics, declare it under metrics in ops.json, and Prometheus picks it up.

TCP connections, queue depth, custom counters
Notifications, emails, payments processed
Alert on any metric from the same Grafana

ops.jsonmetrics block

{
  "service": "notifications-api",
  "type": "internal",
  "metrics": {
    "endpoint": "/metrics",
    "port": 9090,
    "scrape_interval": "15s",
    "labels": [
      "channel",
      "tenant"
    ]
  }
}

Exposed counters and gauges land in Prometheus and become charts and alerts in Grafana automatically.

In-cluster

Stored in-cluster, no egress costs, no external agents.

Logs and metrics live inside your VPC. Nothing is forwarded to a third-party SaaS — so your cloud bill doesn't grow with your observability volume.

Zero egress
Nothing leaves your VPC — no traffic to a third-party data plane.
No external agents
No DataDog / New Relic agents to install, manage or pay for.
Free and open-source
Loki, Prometheus and Grafana are 100% open-source — no per-host fees.

Compared to

How it stacks up against the usual vendors.

Vendor	Pricing model	Egress	Agent
LocalOpsus	Free, open-source	None — in-cluster	No agent
Datadog	Per host + per GB	Egress to Datadog cloud	Agent required
New Relic	Per host + per GB	Egress to New Relic cloud	Agent required
Dynatrace	Per host + DPU	Egress to Dynatrace cloud	OneAgent required
Site24x7	Per monitor + per server	Egress to Site24x7 cloud	Agent required

Vendor pricing models change; check their docs for current details.

Get full-stack monitoring, with your first service.

Free Starter plan, no credit card. Logs and metrics ready the moment your environment is up.

Setting up monitoring, means zero effort.

Open-source observability, wired up for you.

Loki

Prometheus

Grafana