whichdb

Grafana Pyroscope

active

by Grafana Labs · since 2021 · also known as Pyroscope, Phlare

Continuous profiling database that stores CPU and memory profiles over time on object storage.

time-seriescolumnar / olapdatabase engineopen source (copyleft)growing
Website Docs Source★ 11.7K Compare Find similar

Overview

Grafana Pyroscope is a continuous-profiling backend: agents, eBPF collectors or SDKs send CPU, memory, lock and allocation profiles every few seconds, and Pyroscope stores them as time-indexed, deduplicated stack-trace data in object storage so you can compare flame graphs across releases, hosts or time ranges. Grafana Labs acquired the Pyroscope project in 2023 and merged it with its own Phlare database, so the current engine follows the Mimir/Loki architecture with distributors, ingesters and store-gateways. It is AGPL-3.0, integrates with Tempo to link spans to profiles, and is also offered as Grafana Cloud Profiles.

Observability

Strengths & trade-offs

Strengths

  • +Always-on profiling with low overhead via SDKs or eBPF
  • +Flame-graph diffs across time ranges, versions or hosts
  • +Links traces in Tempo to the exact profile of a span
  • +Object storage keeps profile retention inexpensive

Trade-offs

  • −AGPL-3.0 licence
  • −Profiling is a niche signal; most teams adopt it after metrics, logs and traces
  • −Microservices mode shares Mimir-style operational complexity
  • −Language SDK coverage and quality vary

Great for

  • +Finding CPU and memory hot spots in production
  • +Comparing performance between releases
  • +Reducing cloud compute cost

Think twice for

  • −General-purpose time-series storage
  • −Logs or metrics
  • −Ad-hoc SQL analytics

Capabilities

Native or official first-party support · 5 of 25 tracked

✓TTL / expiry
✓Columnar storage
✓Durable persistence
✓Compression
✓Streaming ingest
–Transactions
–Joins
–Secondary indexes
–Full-text search
–Vector search
–Geospatial
–JSON
–Change data capture
–Triggers
–Stored procedures
–Materialized views
–Pub/Sub
–In-memory
–Encryption at rest
–Role-based access
–Time travel
–Branching
–Online schema changes
–Offline sync
–Runs in WebAssembly

Scale & consistency

Comfortable data size
Large · ≤ 1 PB
Horizontal scaling
Native
Multi-region
No
Transactions
None
Consistency
Eventual
Schema
Schemaless
Architecture
Single node, Distributed, Shared storage
Query languages
Label selectors (PromQL-style)

Monolithic for small setups; microservices mode scales components independently over object storage.

Deployment

Self-hosted Vendor-managed cloud Serverless Embedded Edge Kubernetes operator

available on · AWS · Google Cloud · Azure

Grafana Cloud ProfilesGrafana Labs

Ecosystem

compatible withS3
apis & protocolsPyroscope ingest API, pprof push, OTLP profiles, Connect/gRPC query API
client languagesGo, Java, Python, Ruby, .NET, Rust, JavaScript

Alternatives to Grafana Pyroscope

Compare all→
Grafana TempoColumnar / OLAP14.1

Object-storage-backed distributed tracing backend with TraceQL search and trace-derived metrics.

grafana-pyroscope vs grafana-tempo →
SigNozColumnar / OLAP17.0

OpenTelemetry-native APM and observability platform storing logs, metrics and traces in ClickHouse.

grafana-pyroscope vs signoz →
OpenObserveSearch Engine20.4

Single-binary Rust observability platform storing logs, metrics and traces as Parquet on object storage.

grafana-pyroscope vs openobserve →
Grafana MimirTime-Series15.8

Grafana Labs' horizontally scalable, multi-tenant long-term store for Prometheus and OTLP metrics.

grafana-pyroscope vs grafana-mimir →
ParseableColumnar / OLAP18.6

Rust telemetry datalake for logs, metrics and traces on S3, queried with SQL.

grafana-pyroscope vs parseable →
CnosDBTime-Series6.7

Rust cloud-native time-series database with SQL on DataFusion and InfluxDB-style ingestion.

grafana-pyroscope vs cnosdb →

Grafana Pyroscope FAQ

What is Grafana Pyroscope?

Grafana Pyroscope is a time-series database by Grafana Labs, first released in 2021. Grafana Pyroscope is a continuous-profiling backend: agents, eBPF collectors or SDKs send CPU, memory, lock and allocation profiles every few seconds, and Pyroscope stores them as time-indexed, deduplicated stack-trace data in object storage so you can compare flame graphs across releases, hosts or time ranges. Grafana Labs acquired the Pyroscope project in 2023 and merged it with its own Phlare database, so the current engine follows the Mimir/Loki architecture with distributors, ingesters and store-gateways. It is AGPL-3.0, integrates with Tempo to link spans to profiles, and is also offered as Grafana Cloud Profiles.

Is Grafana Pyroscope open source?

Yes. Grafana Pyroscope is open source. License: GNU AGPL v3.0.

Does Grafana Pyroscope support ACID transactions?

No. Grafana Pyroscope does not provide ACID transactions; its consistency model is eventual.

Does Grafana Pyroscope scale horizontally?

Grafana Pyroscope scales horizontally out of the box, with built-in sharding or partitioning. It is comfortable at large scale (≤ 1 PB).

How can I run Grafana Pyroscope?

You can self-host it and use the vendor's managed cloud. Managed offerings include Grafana Cloud Profiles. It is available on AWS, Google Cloud and Azure.

What is Grafana Pyroscope best for?

Grafana Pyroscope is a strong fit for finding CPU and memory hot spots in production; comparing performance between releases; and reducing cloud compute cost. Think twice for general-purpose time-series storage.

What are the best alternatives to Grafana Pyroscope?

Common alternatives to Grafana Pyroscope include Grafana Tempo, SigNoz, OpenObserve, Grafana Mimir and Parseable. The right choice depends on your workload, scale and hosting; WhichDB's smart match ranks them against your requirements.

How popular is Grafana Pyroscope?

Grafana Pyroscope ranks #263 of 844 databases on the WhichDB Score (12.7/100), and #15 among time-series databases. The score combines public adoption signals such as package downloads, GitHub activity, job posts and developer discussion.