Greptime

GreptimeDB v1.0 GA is out. One database for metrics, logs, and traces. SQL + PromQL. Runs on object storage. The big change: Flat SST is now the default format. At 2M series, write throughput is ~4× higher and some query latencies drop up to 10×. Also in v1.0: built-in Perses dashboards (the MCP Server can create them via LLM too), OTLP partial success for traces, and PostgreSQL protocol fixes. 474 PRs from Beta1 to GA. 27 contributors, 8 first-timers. greptime.com/blogs/2026-04-…

@opentelemetry is officially a CNCF graduated project! 🎓🎉 OpenTelemetry has become the trusted de facto observability standard, backed by 12,000+ contributors from 2,800+ organizations and helping teams gain better visibility across distributed systems. Congrats to this incredible community! Read more about the milestone here: bit.ly/4fvcHAb

GreptimeDB now ships an Agent Quickstart prompt on its website and quick start doc — point your AI agent at SKILL.md and it figures out how to install, configure, ingest, and query GreptimeDB. Also biweekly #83: DataFusion 53, pre-cast constants for faster filter pushdown, and new Hotspot + Autopilot dashboard panels. greptime.com/blogs/2026-05-…

DataFusion shipped dynamic filter pushdown upstream last September. GreptimeDB v1.0 wires it into the Mito scan layer. The bound TopK is currently holding gets pushed down to the scan as a runtime predicate — dynamic meaning it tightens as TopK converges, and each row group is evaluated against the latest snapshot, not the one at query start. Side effect: 29s → 0.21s on ORDER BY end_time DESC LIMIT 10 (5B-row traces table). greptime.com/blogs/2026-05-…

One-click copy tells your AI Agent how to install and use GreptimeDB. greptime.com/product/db

IoT Expo 2026 is happening now! Come find us at Booth 115 to chat about IoT data platforms and AI observability. See you there! #IoTExpo #TimeSeries

Added an Agent Onboarding prompt card to our Getting Started docs to help your AI agent install and use GreptimeDB. Covers deploy, configure, ingest, query — MCP-first. Sister skills (pipeline/flow/trigger) hosted alongside. docs.greptime.com/getting-starte…

GreptimeDB v1.0.2 is out. It fixes a query bug where merge_mode + range result cache + OR filters on the time index could return wrong rows, and improves PromQL performance on non-ms time indexes. If you’re on v1.0.0 or v1.0.1, please upgrade github.com/GreptimeTeam/g…

Classic histograms need pre-defined boundaries vs. native histograms auto-scale

Prometheus 3.0's range selectors went from [closed, closed] to (open, closed]. One of several subtle changes backends had to adapt to. Full walkthrough of every notable 3.0 change + the four PRs GreptimeDB shipped to stay drop-in compatible: greptime.com/blogs/2026-05-…

Key features in Prometheus 3.0

We just shipped promql-parser v0.9.0, a PromQL Parser in Rust. Highlights: • declarative refactor for functions • fill*() binop modifiers • mutable visitor trait for AST traversal • limitk and limit_ratio support Details: github.com/GreptimeTeam/p… #PromQL #Rust

GreptimeDB's flat-format queries can now prefilter on any column — tags, fields, timestamps — not just primary keys. While we were at it, mito2 dropped its legacy scan path, about 1,800 lines gone. Biweekly #82: greptime.com/blogs/2026-05-…

Cloudflare Workers run in V8 isolates, not Node.js. Invocations are short-lived, and standard Postgres drivers can't open the connections they expect. D1 / Analytics Engine aren't built for time-series telemetry either. So we wrote up a pattern: GreptimeDB as the Worker's event + metric store. Writes via fetch() (InfluxDB line protocol), reads via Hyperdrive over PG wire. docker compose demo + schema notes: greptime.com/blogs/2026-04-…

Most case studies brag about cost savings to themselves. OceanBase Cloud cut log storage cost 60%+ on GreptimeDB and passed the same 60% straight through to their customers' SQL audit pricing. One year in, they're at 80+ clusters / 300TB / ~1GB/s across 7 public clouds. greptime.com/blogs/2025-07-…

GreptimeDB v1.0.1 is out. Stability and perf fixes on top of 1.0.0. Notable bits: compaction outputs can succeed independently now, staging leader propagates through lease/heartbeat, jieba tokenization is better for zh full-text, mysql metadata limit bumped to mediumblob, manifest recovery scans past the last version. Recommended for 1.0.0 users. github.com/GreptimeTeam/g…

Released GreptimeDB TypeScript Ingester SDK v0.1.0 🎉 Supports gRPC row inserts, streaming inserts, and Arrow Flight bulk writes in one package. @greptime/ingester" target="_blank" rel="nofollow noopener">npmjs.com/package/@grept… Feedback and contributions are welcome! github.com/GreptimeTeam/g…

GreptimeDB biweekly #81: new import-v2 CLI for bulk loading, partition range cache in SeqScan/SeriesScan to cut repeat-scan I/O, and hardened repartition rollback so failed ops don't orphan resources. Four gRPC ingester SDKs also shipped: Go, .NET, Rust (now on crates.io), Erlang. greptime.com/blogs/2026-04-…