←
Greptime961 posts
Greptime avatar

Greptime

@Greptime

The open-source observability database. One columnar engine for metrics, logs & traces on object storage. SQL across signals. PromQL for metrics.

◷ Joined May 2022▧ 282 Photos and videos
Greptime avatar
Greptime@Greptime·Jun 16

GreptimeDB v1.1: you no longer have to lock in a table's partition layout when you create it. Before, only tables created with PARTITION ON COLUMNS could be resharded. Now you can take an existing single-Region table and split it online: ALTER TABLE sensor_readings PARTITION ON COLUMNS (device_id, area) (...) Repartition as data grows, instead of guessing up front. Full blog in detail: greptime.com/blogs/2026-06-1…

Post attachment
Greptime avatar
Greptime@Greptime·8h

Raw output from the run, same data and eval time on both versions:

Post attachment
Greptime avatar
Greptime@Greptime·8h

Run OTel delta counters through counter math and every drop looks like a reset. 12 delta points summing to 1393 gave increase() = 717.8 on GreptimeDB v1.2.1. v1.3 tags delta series, and increase() returns 1393. docs.greptime.com/user-guide…

Post attachment
Greptime avatar
Greptime@Greptime·Nov 17

Kubernetes events disappear after an hour. That's the kube-apiserver default (--event-ttl), so the ImagePullBackOff from last night is gone by morning. New tutorial: one OpenTelemetry Collector with the k8s_objects receiver ships events over OTLP to GreptimeDB. A pipeline extracts namespace, object, reason, and message into columns, drops managedFields, and puts a full-text index on the message. Events stay until the table TTL you set. Then a pod's history is one query: SELECT timestamp, reason, count, note FROM k8s_events WHERE object_name = 'bad-image-66c6f97bd-mtm6v' ORDER BY timestamp; docs.greptime.com/tutorials/… #Kubernetes

Post attachment
Greptime avatar
Greptime@Greptime·Nov 17

GreptimeDB now has a sink connector for Kafka Connect. v0.1.0 is out. If your data already moves through Kafka Connect, getting it into GreptimeDB is a connector config. Two lines decide the table layout: "timestamp.field": "ts", "tag.fields": "device" GreptimeDB creates the table and columns on the first write. Avro, Protobuf, and JSON Schema records get column types from the schema; schemaless JSON works too. Delivery is at-least-once, and a replayed row overwrites the row with the same tags and timestamp, so retries don't leave duplicates. Transient errors are retried, never dropped. Docs: docs.greptime.com/user-guide… Release: github.com/GreptimeTeam/grep…

Greptime avatar
Greptime@Greptime·Nov 16

GreptimeDB v1.3.1 is out. It's a patch release, and worth the upgrade: - Durability: fixes potential data loss during repartition rollback or datanode restart, and periodic local WAL sync (default sync_write = false) fsyncs again - Authorization: selecting a database over HTTP, MySQL, or PostgreSQL now checks permissions; admin functions require ADMIN even through SELECT - Credentials are redacted from diagnostic config and startup logs - Prepared statements on MySQL/PostgreSQL return correct results after partition columns change Before upgrading, read the compatibility notes. Tables with manually configured sparse primary key encoding are validated and not migrated automatically, and MySQL clients without a usable default database must now select an authorized one before running statements that need it. Thanks to 19 contributors, 6 of them first-timers. github.com/GreptimeTeam/grep…

Greptime avatar
Greptime@Greptime·Nov 16

GreptimeDB is good at processing real-time data.

Keep writing demos. Just built Crypto Market Lab to show GreptimeDB in action with Binance trade data: live charts, minute candles, historical replay, and Grafana dashboards. We already have crypto customers using GreptimeDB for market and observability data. Give the demo a try 👇 github.com/GreptimeTeam/demo…

Post attachmentPost attachment
Greptime avatar
Greptime@Greptime·Nov 16

Coming in GreptimeDB v1.4: send InfluxDB Line Protocol, OTLP, and Prometheus Remote Write payloads through PostgreSQL's COPY FROM STDIN, all merged and available in nightly builds. Great work by @Sunng 🚀 Kudos to #PostgreSQL: COPY is a great design. Bulk ingestion goes through the same connection, driver, and auth as any SQL client. 👏

Greptime avatar
Greptime@Greptime·Nov 16

An alert on errors / requests > 0.05 used to fire for a host with no traffic: 0/0 is NaN, and NaN > 0.05 came out true. In GreptimeDB v1.3 it's false, as in Prometheus. Same data, v1.2.1 vs v1.3.1: two hosts fired, now one. Overall, v1.3 passes 94.8% of the PromQL compliance tester, up from 88.7% (labels and values; __name__ not compared). We measured it with promql-compat, an open-source harness that runs the same data through the Prometheus engine and diffs the results. Run it yourself: github.com/GreptimeTeam/prom…

Post attachmentPost attachment
Greptime avatar
Greptime@Greptime·Nov 11

Metrics, logs, and traces in one explore view, with one time range and one set of label filters across all three. Signals Explore ships in the GreptimeDB v1.3 dashboard. Click a trace_id in a log line and the whole trace opens next to it. The live demo runs on the OpenTelemetry official shop Demo: tinyurl.com/4j7f2k8h

Post attachmentPost attachment
Greptime avatar
Greptime@Greptime·Nov 11

Day 1 at #KubeCon Thank you for stopping by and speaking with us about observability data infrastructure. See you tomorrow at booth 262—let’s keep connecting!

Post attachmentPost attachmentPost attachment
Greptime avatar
Greptime@Greptime·Nov 11

High-cardinality telemetry and fast access to historical data are both essential for observability at scale. Our co-founder shares how OpenAI uses GreptimeDB to address these challenges for Spark—and make the data more accessible to AI agents. 👇

#OpenAI gave two talks at Community Over Code EU 2026 on running #Spark at their scale: hundreds of thousands of applications a day. Both end up at GreptimeDB for Spark observability, from opposite ends. One wanted metrics for every Spark task. That's up to tens of millions of new series a day, with logs, metrics and events sitting in separate stores. The other needed the Spark History Server fast enough for agents. Opening a 100 GB+ event log could take up to half an hour, and every agent tool call waited on it. Where it landed: a listener in the Spark driver writes jobs, stages, tasks and executors into #GreptimeDB. The History Server reads from it. The #Flink job that parsed event logs, "probably one of the most expensive Flink apps at OpenAI", is no longer needed. The data sits on S3 as #Parquet, so Spark and Trino query it directly. My read: high cardinality and scattered signals are one problem. They show up together once telemetry is fine-grained enough to be useful. Our notes on both talks, with timestamps: greptime.com/blogs/2026-11-0…

Post attachment