# alibaba/mdrill

for千亿数据即席分析

Repository: https://github.com/alibaba/mdrill
Canonical: https://ross.abutalabs.com/products/mdrill
Homepage: https://github.com/alibaba/mdrill
Language: Java
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2017-11-08T07:51:04+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 4789, "days_push": 3220, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1543, forks 749 (observed 2026-08-28T04:05:01.252547+00:00)

## What it is
Mdrill is an open-source distributed OLAP (online analytical processing) engine from Alibaba's AdMom team, built in Java on top of JStorm/Hadoop-style infrastructure. It uses columnar storage, inverted indexes, compression, caching, and partitioning to deliver second-level ad-hoc queries over hundreds of billions of rows on as few as 10 commodity machines, with both offline batch and Kafka-based realtime ingestion.

## Use cases
- run second-level ad-hoc analytical queries over hundreds of billions of rows
- analyze TB-scale data on a small cluster of commodity machines instead of an expensive data warehouse
- ingest billions of rows per day in realtime via Kafka
- do incremental partition-based updates of offline data
- slice and filter data across arbitrary combinations of 80-400 dimensions
- build a self-service big data analytics platform (like Alibaba's AdHoc)

## When to choose
- you need low-cost, sub-minute ad-hoc OLAP queries over very large fact tables with many dimensions
- you want both realtime streaming ingestion and offline batch loading in one system
- you are running on a small on-premise cluster and cannot afford commercial distributed warehouses

## When to avoid
- you need a maintained project - the repository was archived in October 2025 and the last release was in 2017
- your workload fits modern alternatives like ClickHouse, Druid, or Doris which have larger communities
- you need cloud-native deployment, rich SQL coverage, or an active ecosystem

## Facets
- artifact type: service
- maturity: abandoned
- function: database, search-engine, streaming, etl
- domain: big-data, analytics, databases
- platform: jvm, self-hosted
- tags: olap, adhoc-analytics, columnar-storage, inverted-index, distributed-query, realtime-ingestion, sql, archived, data-engineering, linux, docker

## Member repositories
- alibaba/mdrill (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:01.252547+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:30:40.371280+00:00, confidence not recorded.
  - readme: https://github.com/alibaba/mdrill (fetched 2026-08-28T04:05:01.252547+00:00, sha d466491a0589)
  - homepage: https://github.com/alibaba/mdrill (fetched 2026-08-29T11:31:52.768069+00:00, sha 08e531bc8173)
- Data as of 2026-08-30T08:39:29.467469+00:00.