1.2 KiB
1.2 KiB
title | author | abstract | slide_level | theme | date |
---|---|---|---|---|---|
Druid pour l'analyse de données en temps réel | Yann Esposito | Druid expliqué rapidement, pourquoi, comment. | 2 | solarized | 7 Avril 2016 |
Intro
Plan
- Introduction ; pourquoi ?
- Comment ?
Expérience
- Real Time Social Media Analytics
Demande
- Twitter:
20k msg/s
,1msg = 10ko
pendant 24h - Facebook public: 1000 à 2000 msg/s en continu
En pratique
- Twitter: 400 msg/s en continu, pics à 1500
Origine (PHP)
Introduction
- Traitement de donnée gros volume + faible latence
- Typiquement
pulse
Druid
Who
Metamarkets
Goal
Druid is an open source store designed for real-time exploratory analytics on large data sets.
hosted dashboard that would allow users to arbitrarily explore and visualize event streams.
Concepts
- Column-oriented storage layout
- distributed, shared-nothing architecture
- advanced indexing structure
Features
- fast aggregations
- flexible filters
- low latency data ingestion
arbitrary exploration of billion-row tables tables with sub-second latencies