Metrics and monitoring

Metrics without cardinality limits

When Prometheus crashes under a cardinality spike, you lose visibility at the worst possible moment. Uptrace stores metrics in a columnar engine with no per-series limits. Query with the PromQL you already know, get pre-built dashboards automatically, and alert on anything.

Start collecting metrics Schedule a demo

14-day free trial · No credit card required

Instant visibility from the first metric

The moment you start sending metrics, Uptrace shows you an interactive explorer and pre-built dashboards for your stack, no configuration required

Metric Explorer: browse, filter, and chart any metric interactively
Pre-built dashboards for host metrics, Docker, PostgreSQL, MySQL, Redis, Elasticsearch
Dashboards appear automatically when matching metrics are detected
See all received metrics organised by service and instrument type
Identify coverage gaps before they become incidents

No more cardinality limits

Prometheus degrades when cardinality spikes. Uptrace uses a columnar store that handles millions of unique label combinations, and you query it with the PromQL you already know

Columnar storage handles millions of unique label combinations
No per-series pricing penalty: add any label dimension freely
PromQL-compatible: existing queries, recording rules, and alerts work as-is
OpenTelemetry histograms stored natively for accurate p50/p95/p99
Up to 95% storage reduction via ClickHouse compression

Prometheus Uptrace

Collect from any source

Uptrace accepts metrics from OpenTelemetry SDKs, Prometheus remote write, and all major collector pipelines, no re-instrumentation required

Native OTLP gRPC and HTTP ingestion
Prometheus remote write: redirect existing Prometheus without pipeline changes
StatsD via the OpenTelemetry Collector
AWS CloudWatch metrics via native collector receiver
Horizontal auto-scaling handles millions of metric events per second

Custom dashboards for any stack

Beyond pre-built templates, Uptrace lets you build parameterised dashboards with any metric: grid layouts, table views, and YAML-defined templates you can version-control

Custom grid dashboards with collapsible row groups and multiple chart types
Table dashboards for aggregation views per label dimension
Template variables as dynamic dropdown filters at the top of every dashboard
Variable substitution to reuse dashboard layouts across environments
YAML-based dashboard templates for version-controlled definitions

Alert before users notice

Create threshold monitors on any metric and get notified via Slack, PagerDuty, Teams, or webhook, before users notice something is wrong

Metric monitors with configurable thresholds and evaluation windows
Alert on host CPU, memory, disk, and network from hostmetrics directly
Multi-condition monitors combining metric thresholds with span and log rules
Channels: Slack, Teams, PagerDuty, Opsgenie, Telegram, webhook
Silence windows for planned maintenance

Better together

Each signal is more powerful when correlated with the others. Uptrace stores traces, logs, and metrics in one place, no tool-switching, no data gaps.

Distributed tracing

Find the exact span that broke your request, across every service and database call.

Log management

Search structured logs at any scale and jump from any span to its logs in one click.

Common questions, clear answers

Is Uptrace a replacement for Prometheus?

Uptrace can fully replace Prometheus as a metrics storage backend: it accepts Prometheus remote write and supports PromQL. It also adds traces, logs, and alerting in the same platform, so you no longer need a separate Grafana installation to visualize the data.

Can I use my existing PromQL queries?

Yes. Uptrace supports the full PromQL query language plus extensions for aliases, joins, and cross-metric expressions. Recording rules and alerting rules written for Prometheus work as-is.

How does Uptrace handle high cardinality better than Prometheus?

Prometheus stores metrics in a time-series database optimised for low cardinality. When the number of unique label combinations grows, memory and query time degrade significantly. Uptrace uses ClickHouse, a columnar store, which is designed for exactly this kind of high-cardinality analytical workload.

How does Uptrace handle OpenTelemetry histograms?

Uptrace stores OpenTelemetry histograms natively and computes accurate percentiles (p50, p75, p90, p95, p99) without the approximation errors that come from converting histograms to Prometheus summary format.

How do I migrate from Prometheus and Grafana?

Point your Prometheus remote write endpoint at Uptrace and metrics start flowing immediately. Existing PromQL queries work without changes. Pre-built dashboards for common stacks appear automatically once Uptrace detects the matching metrics.

Ready to see every metric in context?

Deploy in minutes with Docker. Open source. Fixed pricing. Your data stays on your infrastructure.

Start collecting metrics

14-day free trial · No credit card required

Learn more

Documentation Pricing Schedule a demo