Greptime

🧵We published 93 blog posts in 2025. One article got over half our traffic. Here's what we learned about tech content 👇 1. The runaway winner: "Observability 2.0 and the Database for It" Not a product launch. An opinion piece on where o11y is heading. Timing + a clear take = viral (for a database blog). 2. What else worked: 📊 Benchmarks—4 of top 15 were perf comparisons 🦀 Rust deep-dives—Android edge cases, source reading 👤 People stories—contributor interviews cracked top 10 3. The brutal math: Top 1 post = 50%+ of traffic Top 5 posts = ~70% of traffic Power law is real. Most posts get little reach. A few carry everything. 4. Lesson: depth > volume. One well-timed, technically solid post beats 10 mediocre ones. Full breakdown: greptime.com/blogs/2026-01-… #Greptime

Actually, repartitioning is not just for dealing with hot partitions or cold fragments; it can also be used to scale a table's read and write capacity by adding more regions.

Repartition — one of the most important features in GreptimeDB v1.0-rc — lets you split and merge partitions online without physical data migration. Hot partition? Split it. Cold fragments piling up? Merge them back. We wrote a practical guide with step-by-step SQL examples covering both Mito Engine and Metric Engine tables, plus checklists for running this safely in production. Read it: greptime.com/blogs/2026-03-…

This made our day. 🙌 SDK → GreptimeDB — that's the architecture we designed for. Thanks hebo.ai for the detailed writeup, especially the Langfuse comparison and the OTel GenAI conventions deep dive.

Greptime Biweekly #78 🎉 6,000 GitHub stars! 🌟 Thanks to everyone who has starred, contributed, or used the project.♥️ Plus: new anomaly detection functions (zscore, iqr, mad), PostgreSQL COPY TO STDOUT for ADBC, admin GC commands for tables/regions, and DataFusion upgrade. Read more: greptime.com/blogs/2026-03-…

We just released GreptimeDB v1.0.0-rc.2, continuing our march toward v1.0 GA. Key highlights: - 100×+ faster TopK queries — ORDER BY ... LIMIT 10 improved from ~28.86s to ~0.21s via dynamic filter pushdown 🚀 - PostgreSQL COPY TO STDOUT support, enabling the ADBC-postgres optimized export path - Prometheus 3.x selector/lookback behavior alignment - Critical repartition stability fixes - New anomaly detection functions: anomaly_score_zscore, anomaly_score_iqr, anomaly_score_mad 84 PRs merged from 13 contributors👏. We strongly recommend upgrading. Blog: greptime.com/blogs/2026-03-… Release notes: github.com/GreptimeTeam/g… #GreptimeDB #OpenSource

LLM app breaks in production. You open Jaeger for traces, Prometheus for token metrics, Elasticsearch for the actual conversation. Three tabs, one trace_id you're copy-pasting between them. There are vendor-specific LLM observability platforms. This demo takes a different path: open standards (OTel GenAI) + open-source database (GreptimeDB). One SQL query joins traces with conversations. The whole thing runs with docker compose. Blog post: greptime.com/blogs/2026-03-… #GenAI #OpenTelemetry

Most observability databases need a separate admin API for schema changes, TTL, indexes and metadata inspection. GreptimeDB does it all in SQL.🧩 DROP/ALTER TABLE, SET TTL, CREATE INDEX, SHOW CREATE TABLE — even INFORMATION_SCHEMA for inspecting metadata. No custom control plane, no context switching. Docs: 📄 docs.greptime.com/user-guide/man… 📄 docs.greptime.com/reference/sql/… #Observability #OpenSource #SQL

With dynamic filter pushdown into the scan layer, queries on non-time-index columns can prune irrelevant data much earlier. For ORDER BY end_time DESC LIMIT 10, this delivered 700x faster sort compute and cut end-to-end latency from 28.86s to 0.21s. github.com/GreptimeTeam/g… #Performance #OpenSource

Keep improving the performance and memory efficiency of Prometheus Remote Write request decoding: 👉 Replaced generic UTF-8 validation for Prometheus label names with a specialized byte-level lookup table and loop unrolling 👉 Reduced memory allocations by replacing Bytes cloning with &'static [u8] for label names Resulted in a ~10% performance gain! github.com/GreptimeTeam/g… #OpenSource #RustLang

Benchmarked the GreptimeDB .NET Ingester SDK against GreptimeDB v1.0 RC1 on Apple M1 Max (~70 bytes/row, 9 columns, 100K rows/task): - gRPC Unary: ~960K rows/s - gRPC Streaming: ~1.3M rows/s - Arrow Flight: ~2M rows/s (~140 MB/s) All examples are single-file .NET 10 scripts — just dotnet run benchmark.cs. github.com/GreptimeTeam/g…

Announcing GreptimeDB .NET Ingester v0.1.0 — the official .NET SDK for writing observability & IoT data to #GreptimeDB! Features: - gRPC unary & streaming writes - Apache Arrow Flight bulk ingestion - .NET 6/7/8/9/10 support - DI integration out of the box dotnet add package GreptimeDB.Ingester Github: github.com/GreptimeTeam/g… NuGet: nuget.org/packages/Grept… #Dotnet #GreptimeDB #OpenSource

We'll release the #GreptimeDB v1.0 RC2 version later this week. It mainly includes fixes for a few bugs in region repartition, along with some performance optimizations.

We partnered with @EMQTech to build EMQX Tables — GreptimeDB running natively inside EMQX Cloud as a fully managed time-series database. If you're running MQTT, your device telemetry now goes from broker to storage to SQL query without provisioning a separate DB or wiring up ingestion pipelines. What makes it work: → Object storage-first (S3/GCS) — 3-20x cheaper for high-volume IoT data → Schemaless ingestion — no migrations when new device types show up → Native InfluxDB Line Protocol, Prometheus Remote Write, and OTLP support Blog → greptime.com/blogs/emqx-gre… Try it → emqx.com/en/cloud/emqx-… #TimeSeries #EMQX #GreptimeDB

We've fully refreshed the GreptimeDB Concepts docs. If you're interested in GreptimeDB's design principles, data model, and key features, these documents are a good place to start: docs.greptime.com/user-guide/con… #GreptimeDB #OpenSource

GreptimeDB Biweekly Update No. 76 & 77 🔧 We're combining four weeks of progress in this update. Here's what shipped: Performance & Compatibility • BulkMemtable lazy encoding cuts L0 SST files by ~38% in TSBS workloads • PostgreSQL protocol now uses arrow-pg with proper unsigned integer handling • Prometheus 3.x selector and lookback behavior fully supported • Partition staging flow fixed for safer repartitioning Big Announcements 📢 v1.0 RC1 is here — we're approaching the 1.0 milestone 📍 Our 2026 Roadmap is live — check out what's ahead for GreptimeDB this year Community Over the past month, 58 PRs merged with contributions from linyihai, YZL0v3ZZ (welcome!), and ZonaHex. Thanks to everyone building with us. We also redesigned greptime.com with a cleaner look. Take a look and let us know what you think. For detailed technical changes and good first issues, read the full report: greptime.com/blogs/2026-02-… #OpenSource #Database #GreptimeBiweekly #Observability #GreptimeDB

That's cool! Thanks for sharing it!