Everyone can contribute! Let's learn together in a weekly cafe ☕

General

• Agenda
• Guests: Vlad Ilyushchenko, Nicolas Hourcard, David G. Simmons, Niclas Mietz, Michael Friedrich, Michael Aigner, Nico Meisenzahl
• Next ☕ chat #2

Highlights

• QuestDB introduction, a high-performance Time Series Database
• Live demo
• Charts, clustering, memory management, scaling, roadmap

Recording

Enjoy the session!

Insights

• Introduction into QuestDB: Column based, time partitioned, append only (immutable), repeatable read consistency, highly optimized
• Integrations
  • InfluxDB line protocol v1 or v2 (are there changes)
  • Prometheus support directly?
• Authentication support?
  • Certificate based auth
• Development and Contributions
  • Build from source, Java as dependency: https://github.com/questdb/questdb#building-from-source
  • Docker dev containers after CI/CD is fully implemented
  • Slack channels: https://github.com/questdb/questdb#support--contact
• Demo: http://try.questdb.io:9000/

• Grid and chart views
  • Idea: Have PNG/PDF export
  • Embed charts into GitLab alerts and incidents
• Code base: Java 11 primarily. C is only for IO.
  • Query and hashmap lookups, insights in Agner Fog, vector class 2 and efficient hashmaps in C++
• Clustering with QuestDB
  • Column first ingestion and inserting simultaneously mechanism to be built
• Roadmap for next year
  • Replication (for OSS)
  • Out of order. Currently the series needs to be inserted in order, this is good for historical metrics replay.
  • Roadmap on GitHub
• Enterprise version
  • First year focus on OSS, Enterprise is planned at ~18 months
  • Pulsar as Open Core model.
  • Uses QuestDB as library.
  • Superlarge environments at scale.
  • UDP based replication for nodes closer together.

---

Date published: September 23, 2020