Polartrace

Product

Everything that ships today. Nothing that doesn't.

Six product surfaces, one signal graph: APM, distributed tracing, logs, monitors, runtime metrics and Kubernetes - for Node.js and Python teams.

Start freeRead the docs

APM

The whole fleet on one dashboard.

Requests, error rates, latency percentiles and active alerts for every service - with per-service drill-down into performance, endpoints, database operations and resource usage.

  • Service list with p90/p99 latency, error rate and trends
  • Endpoint table ranked by total time, one click from its traces
  • Top database operations and per-service dependency graph
  • Process CPU, memory and event-loop lag per service
Polartrace dashboard: attention cards for payment gateway, checkout and search alerts, fleet health, requests and errors, with the checkout critical alert and fleet health cards floating out of the window

Distributed tracing

One request, every hop, all the timings.

OpenTelemetry-powered spans with automatic context propagation. HTTP calls, MongoDB and PostgreSQL queries, Redis commands - laid out on a waterfall you can read at 2 a.m.

  • Span waterfall with kind, status, attributes, events and links
  • Latency percentile lines: p50, p75, p90, p95, p99, p99.9, max
  • Log-scale duration histogram with percentile markers
  • Search by service, span name, HTTP method or trace ID
Polartrace trace view for POST /api/reports/kpis: span waterfall, correlated logs and span attributes, with the request timeseries and percentile chart floating below

Log management

Request logs that carry their own context.

Every request arrives with method, path, status, duration, headers and - for Node.js - captured console output, error stacks and the trace ID that joins it to its distributed trace.

  • Faceted explorer: filter by service, level, status class, path and time
  • Endpoint patterns with error counts and sparklines
  • Live tail that updates every few seconds
  • One click from a failing request to its trace (Node.js)
Polartrace logs explorer: services and log level facets, log volume chart and log rows, with a selected GET /api/health record and its attributes floating out

Monitors & alerts

Thresholds with a warning shot.

Monitors evaluate every minute across error rate, p99 latency, CPU, memory, request volume and Kubernetes health - with separate warning and critical thresholds, and alerts that tell you when things recover, too.

  • 8 monitor metrics including Kubernetes pod & node health
  • Dual warning + critical thresholds with live preview
  • Slack and email delivery with a trend chart and deep link
  • Edge-triggered: one notification per state change, no spam
Polartrace monitor detail for high latency on search-api: alert timeline, event evidence chart with breach window, alert details and Slack notification floating out

Runtime & host metrics

The metrics that explain your latency.

Percentiles tell you something is slow; runtime metrics tell you why. Process CPU, memory RSS, host load and Node.js event-loop lag are sampled continuously by the same agent that traces your requests.

  • p50–p99.9 latency per service and per endpoint
  • Process CPU, memory and host load averages
  • Event-loop delay mean and p99 for Node.js
  • Container-aware: cgroup CPU and memory limits included
Polartrace resource usage for session-store: CPU, memory RSS, event loop lag and instance cards, process CPU chart with a hover tooltip, memory breakdown chart

Kubernetes

Cluster health beside your services.

The cluster agent snapshots nodes, pods, containers, workloads and HPA state, streams Kubernetes events with warning reasons, and can collect container logs for namespaces you opt in. Setup is guided from the console's integration wizard.

  • Overview, clusters, topology, nodes, pods, events and logs tabs
  • Restarts, pressure conditions, capacity vs usage
  • Warning events grouped by reason
  • Read-only agent, single deployment, guided install
Polartrace Kubernetes topology for prod-apac-large: cluster, nodes and pods graph with the cluster status panel and an auth-service pod detail panel floating out

Governance

Grown-up controls without the enterprise sales call.

Environments

Separate production from staging and development - same service names, isolated telemetry.

Teams & members

Group services by team. Invite unlimited team members on every plan - users are never billed.

Custom-role RBAC

Compose roles from 25 permission keys across APM, observability, monitoring, billing and IAM. Pro and above.

Audit log

Every monitor edit, key rotation and invite recorded - 30-day retention on Pro, 395-day on Enterprise.

Plus sign-in with Google, login alerts for new devices, and client-side redaction of secrets before telemetry ever leaves your servers.

Free plan: 5 GB/month - never billed

Start monitoring in minutes.

Two commands to your first trace. Start on the free plan - no credit card required.

Start freeRead the docs