dennis zhuang
0d2f1ff6cd
I hereby agree to the terms of the [GreptimeDB CLA](https://github.com/GreptimeTeam/.github/blob/main/CLA.md). ## Refer to a related PR or issue link (optional) Targets `narthex`. Stacked on #9684 for `merge_same_labelset` and the golden fixture helpers; review only the two commits `feat(promql): evaluate classic histogram_quantile and histogram_fraction over grid series` and `feat(promql): evaluate native histogram functions over grid series`. ## What's changed and what's your intention? ### Problem The native path has no kernels for the histogram functions: `histogram_quantile` and `histogram_fraction` over classic (`le`-bucketed) and native histograms, and `histogram_count`, `histogram_sum`, `histogram_avg`, `histogram_stddev` and `histogram_stdvar` over native histograms. ### Change `histogram_eval::evaluate_histograms` takes the evaluated input series and a `HistogramFunction` (`Quantile(q)` or `Fraction { lower, upper }`, constants as in the ordinary planner) and follows Prometheus `funcHistogramQuantile`, `funcHistogramFraction` and `resetHistograms`: - Float series with the same labels other than `le` form a classic histogram; a series whose `le` does not parse is ignored. Parsing follows Go's `strconv.ParseFloat`: finite literals out of the f64 range (`1e400`) and a signed NaN are rejected, and `_` between digits is accepted. - Native histogram samples are evaluated one by one. At a point where a native histogram and classic buckets of the same labels (`le` aside) are both present, neither is evaluated. - Each field is evaluated on its own. Results drop `le`, `__name__`, `__type__` and `__unit__`; series that then share a label set go through `merge_same_labelset`. Classic buckets (`histogram_eval::classic`) port `BucketQuantile` and `BucketFraction`. At each point the present buckets are sorted with `go_sort::sort_by` and Go's `<` comparator, so +0/-0 and NaN bounds keep Go's order, and buckets with equal bounds are coalesced. For quantiles, counts within a relative 1e-12 of the previous count or below it are raised to it. The quantile bucket is found with Go's `sort.Search` bisection, so NaN counts select the same bucket. Native histograms (`histogram_eval::native`) port `HistogramQuantile`, `HistogramFraction` and `histogramVariance`. The port includes: - the bucket order and zero-bucket clipping of `AllBucketIterator`/`AllReverseBucketIterator`; - bucket bounds `2^(idx / 2^schema)` from the standard library, with `f64::MAX` for the last bucket of finite values as in Go's `getBound`; This covers the reverse iteration for `q >= 0.5`, the count clamp, custom-bucket edge buckets, and the trailing walk that Prometheus runs for NaN observations, which leaves the last bucket in place for interpolation. `histogram_eval::map_native_histogram` applies `NativeHistogramFunction::{Count, Sum, Avg, StdDev, StdVar}` to the histogram samples of one series, dropping float samples and `__name__`; colliding label sets are left to `merge_same_labelset`, as for other per-series kernels. Review focus: - `classic::bucket_quantile`: the `q` checks come before the `+Inf` check, the `b == 0 && upper_bound <= 0` branch, and the interpolation `start + (end - start) * (rank / count)`. - `classic::ensure_monotonic`: `prev` is not advanced when a count is raised. - `native::all_buckets` and `native::bound`: iteration order, clipping to the zero threshold, and the last finite bucket (`f64::MAX`). - `native::quantile`: the last visited bucket (empty or not) is the one interpolated, including after the NaN-observation walk. - `evaluate_histograms`: the classic/native conflict is decided per point and field. ### Non-goals - Native histogram `sum`/`avg` aggregation, rate functions and arithmetic. - Non-constant `q`, `lower` and `upper`. - Planner wiring. ### Constraints - The ordinary `HistogramFold::evaluate_row` differs from Prometheus for classic histograms: it errors when the last bucket is not `+Inf`, has no small-delta tolerance, replaces non-finite counts with the previous count, and interpolates as `lower + (upper - lower) / count * rank`. The ordinary native quantile (`common_query::native_histogram::quantile_histogram_view`) has no reverse iteration for `q >= 0.5` and derives bounds with `powf` instead of Go's table. This PR follows Prometheus 3.15. - `le` values in Go-only float syntax (hex floats such as `0x1p-2`) do not parse with Rust's `f64::from_str` and are ignored. ### Verification - `cargo nextest run -p promql`: 318 passed. - `cargo clippy -p promql --all-targets --all-features -- -D warnings`, `cargo fmt --check`, `hawkeye check`: clean. - `classic_histograms_match_prometheus`: 300 fixtures produced by the Prometheus v3.15.0 engine built for amd64 (`histogram_eval/classic_cases.txt`). - Bucket layouts: non-positive bounds, a bound written twice (`1`/`1.0`), no `+Inf`, only `+Inf`, unparsable, out-of-range or missing `le`, and the `inf` spelling. - Counts: monotonic, non-monotonic, around the 1e-12 tolerance, NaN, `+Inf`, negative, zero and subnormal. - Inputs are passed in reverse order. 32 cases expect the duplicate-labelset error. - `native_histograms_match_prometheus`: 420 fixtures (`histogram_eval/native_cases.txt`, 60 per function). - Histograms: schemas -4, -1, 0, 1, 3, 8 and custom buckets, with gaps, zero-length spans, empty buckets, zero thresholds from 0 to 0.5, NaN sums with uncounted observations, and counts below the bucket total. - Inputs mix floats and histograms, and include classic buckets of the same labels. - Inputs are read back from the engine, so they carry its span layout. - Values in both fixtures compare within promql-compat's relative tolerance (1e-10, 1e-6 for `histogram_stddev`/`histogram_stdvar`), NaN as any NaN. - Mutation check, classic: each of these fails a test: - no coalescing (quantile or fraction), no forced monotonicity, no small-delta tolerance, an absolute tolerance at zero; - a linear search, no `+Inf` check, `q` checks after it, no non-positive first-bucket branch, the ordinary interpolation order; - accepting `1e400`, keeping `__name__`, a zero lower bound for a non-positive first fraction bucket, interpolating inside a `-Inf`-wide bucket, no merge, sorting with NaN bounds last. - Mutation check, native: each of these fails a test: - always iterating forward, no NaN-observation walk, no zero-threshold clipping, no clamp of the count; - shifted bucket bounds, no `f64::MAX` for the last finite bucket; - a plain variance sum, a geometric midpoint for custom buckets; - the histogram count as the fraction total under NaN observations, linear interpolation in exponential buckets; - no custom `-Inf` branch, no zero-bucket side adjustment; - evaluating conflicting classic and native histograms, keeping `__name__`, no zero bucket in reverse order. - `histogram_functions_drop_metadata_labels`: `histogram_quantile` drops all metadata labels, `histogram_count` only `__name__`. - `classic_bounds_keep_the_first_signed_zero`: bounds +0 and -0 compare equal and keep input order, so `1 / histogram_quantile(0, ...)` over `le` "+Inf", "0", "-0" is +Inf. ## PR Checklist Please convert it to a draft if some of the following conditions are not met. - [x] I have written the necessary rustdoc comments. - [x] I have added the necessary unit tests and integration tests. - [ ] This PR requires documentation updates. - [x] API changes are backward compatible. - [x] Schema or data changes are backward compatible. - [ ] This PR needs to be backported to release branches, and I have added the `backport-<target>` labels (e.g. `backport-v1.3` targets `release/v1.3`).
Metrics, logs, and traces.
One engine, on your infrastructure.
A columnar database for metrics, logs, and traces on object storage. Apache-2.0 licensed core.
User Guide · API Docs · Roadmap 2026 · Slack
stable for production · canary includes pre-releases · nightly is a weekly snapshot of main
Introduction
GreptimeDB is an open-source observability database. Metrics, logs, and traces run on one columnar engine over object storage and share one table model: tags, timestamp, and fields. When signals carry common identifiers such as service, host, or trace ID, you can correlate them in SQL without moving data between databases.
Ingest through OpenTelemetry, Prometheus Remote Write, Loki Push, or Elasticsearch Bulk. Use SQL across observability data and PromQL for metrics. Migrate ingestion one signal at a time without rebuilding your collectors.
One Query Across Signals
OpenTelemetry ingestion writes spans to opentelemetry_traces and log records to
opentelemetry_logs. Both tables carry trace_id, so correlating them is a join:
-- The slowest failed spans in the last hour,
-- with the log lines emitted inside those same traces.
SELECT
t.service_name,
t.span_name,
t.duration_nano / 1000000 AS duration_ms,
l.timestamp AS log_time,
l.severity_text,
l.body
FROM opentelemetry_traces t
JOIN opentelemetry_logs l ON l.trace_id = t.trace_id
WHERE t.timestamp > now() - INTERVAL '1' HOUR
AND t.span_status_code = 'STATUS_CODE_ERROR'
ORDER BY t.duration_nano DESC
LIMIT 20;
Metrics join the same way, on any tag the tables share, such as service,
host, or pod.
Why You Might Use It
- You run Prometheus plus Loki or Elasticsearch and want one backend instead of three
- You have outgrown Prometheus on cardinality or retention and don't want the Thanos/Mimir operational surface
- You are hitting Loki's query performance limits as log volume grows
- You need long retention on object storage without a separate analytics stack
- You want to query telemetry with SQL, not only a domain query language
- You are storing GenAI or agent telemetry (OTel GenAI conventions) alongside infrastructure signals
Learn more in Why GreptimeDB.
What's Supported
| Ingest | OpenTelemetry (OTLP), Prometheus Remote Write, Loki Push, Elasticsearch Bulk, InfluxDB line protocol, gRPC |
| Query | SQL, PromQL, Jaeger-compatible trace queries, MySQL and PostgreSQL wire protocols |
| Storage | S3, GCS, Azure Blob and S3-compatible endpoints as primary storage, with memory and local-disk caches |
| Built in | Retention policies, downsampling, continuous aggregation, explicit table partitioning, and inverted / skipping / fulltext indexes |
Compute and storage are disaggregated: object storage holds the data, while memory and local-disk caches keep recent and frequently queried data close to compute.
Benchmarks
- Agent RCA Bench: LLM agents doing root cause analysis over GreptimeDB versus Prometheus + Loki + Tempo. 40% fewer wrong diagnoses, 48% fewer input tokens, 45% lower cost (write-up)
- GreptimeDB tops JSONBench's billion-record cold run test
- TSBS Benchmark
- More benchmark reports
Compatibility and Migration
Compatibility is per protocol, and query-side coverage is narrower than ingestion.
| Compatible | Not compatible | |
|---|---|---|
| Prometheus | Remote Write ingestion; PromQL queries | Gaps are listed in PromQL compatibility |
| Loki | Push ingestion; dual-write through Grafana Alloy makes the cutover gradual | LogQL and the rest of the Loki query API |
| Elasticsearch | _bulk ingestion in the open-source core; QueryDSL partially, in Enterprise |
Most other Elasticsearch APIs |
Limitations and Edition Boundary
Cluster deployment, object storage, the Flow engine, and every ingestion protocol listed above are in the Apache-2.0 build. Repartitioning, region migration, and index creation are manual operations there.
Read replicas, workload isolation, and automated repartitioning are GreptimeDB Enterprise features, along with enterprise security and governance. The Enterprise overview has the current list, and pricing has the edition comparison.
Architecture
GreptimeDB can run in two modes:
- Standalone — single binary for development and small deployments.
- Distributed — four components, each independently scalable:
- Frontend — protocol entry (OTel, Prometheus, MySQL/PostgreSQL, gRPC, ingestion APIs for Elasticsearch/InfluxDB/Loki) and the distributed query engine. Stateless, scales horizontally.
- Datanode — region engine with WAL, memtable, SST, cache, compaction, and indexes. Persists data to object storage. Elastic.
- Metasrv — metadata, routing, repartitioning, and security. Backed by a pluggable KV layer (etcd or RDS).
- Flownode (optional) — continuous flow computation (streaming and materialized views).
For deeper coverage, see the architecture doc or DeepWiki.
Try GreptimeDB
For AI agents — paste this prompt into your agent:
Read https://docs.greptime.com/SKILL.md and follow the instructions
to deploy, configure, ingest, and query GreptimeDB.
docker run -p 127.0.0.1:4000-4003:4000-4003 \
-v "$(pwd)/greptimedb_data:/greptimedb_data" \
--name greptime --rm \
greptime/greptimedb:latest standalone start \
--http-addr 0.0.0.0:4000 \
--grpc-bind-addr 0.0.0.0:4001 \
--mysql-addr 0.0.0.0:4002 \
--postgres-addr 0.0.0.0:4003
Dashboard: http://localhost:4000/dashboard
Read more in the full Install Guide.
Troubleshooting:
- Cannot connect to the database? Ensure that ports
4000,4001,4002, and4003are not blocked by a firewall or used by other services. - Failed to start? Check the container logs with
docker logs greptimefor further details.
Getting Started
Build From Source
Prerequisites:
- Rust toolchain — stable, pinned by
rust-toolchain.toml - Protobuf compiler (>= 3.15)
- C/C++ building essentials:
gcc/g++/autoconfand the glibc dev package (libc6-devon Ubuntu,glibc-develon Fedora) - Python toolchain (optional, only for some test scripts)
Build and run:
make # build greptime binary
cargo run -- standalone start # start in standalone mode
Common dev commands:
make fmt # format Rust code
make clippy # lint (fails on warnings)
make test # unit + integration tests (uses cargo-nextest)
make sqlness-test # SQL regression tests
See the Contribution Guidelines for the full developer workflow.
Tools & Extensions
- Kubernetes: GreptimeDB Operator
- Helm Charts: Greptime Helm Charts
- Dashboard: Web UI
- gRPC Ingester: Go, Java, C++, Erlang, Rust, .NET, TypeScript
- Grafana Data Source: GreptimeDB Grafana data source plugin
- Grafana Dashboard: Official Dashboard for monitoring
Project Status
GreptimeDB is generally available, with stable APIs and regular releases. It runs in production at scale — OceanBase Cloud operates 80+ GreptimeDB clusters managing 300 TB of logs, cutting log storage cost by 60%+ after migrating from Grafana Loki. See more in case studies.
Release lines and support windows are in the version reference. For where the project is going, read the v1.0 highlights and the 2026 roadmap.
Community
We invite you to engage and contribute!
If GreptimeDB is useful to you, please star the repo.
License
GreptimeDB is an open-core project. Its core is licensed under the Apache License 2.0.
A small set of peripheral, enterprise-only features are gated behind the
enterprise Cargo feature (not built by default) and are governed by the
separate GreptimeDB Enterprise License. Source files under
that license carry an explicit Enterprise License header.
Commercial Support
Scaling observability on your infrastructure? GreptimeDB Enterprise adds the operational, security, and support layer for production deployments. Contact us for details.
Contributing
- Read our Contribution Guidelines.
- Explore Internal Concepts and DeepWiki.
- Pick up a good first issue and join the #contributors Slack channel.
Acknowledgement
Special thanks to all contributors! See AUTHOR.md.
- Uses Apache Arrow™ (memory model)
- Apache Parquet™ (file storage)
- Apache DataFusion™ (query engine)
- Apache OpenDAL™ (data access abstraction)
All trademarks, logos, and brand names referenced in this README and in the Overview diagram are the property of their respective owners. Their use is for identification purposes only and does not imply endorsement or affiliation.