Marija Selakovic
summary
Claims that CrateDB handles industrial time series by combining SQL with Lucene-based indexing and columnar doc values for fast aggregations, using a shared‑nothing distributed architecture with per-shard replication and automatic failover to keep data available. It stores nested device data as JSON objects, supports arrays to pack many measurements per row, offers joins (nested‑loop and distributed hash) and JavaScript user‑defined functions, and makes newly ingested rows queryable almost instantly. Cloud signup includes 200 free credits.
AI-written summary · about
chapters (24)
- 0:00Introduction
- 1:16CrateDB and speaker
- 2:00Agenda and time series
- 3:04What is CrateDB
- 6:14Industrial time series
- 8:04Storing industrial time series
- 9:14Time series solutions
- 11:45Industrial time series challenges
- 12:26Availability in manufacturing
- 13:29CrateDB high availability
- 16:58Logical replication
- 19:07Semi-structured industrial data
- 20:35Hierarchical time series
- 21:18Document store feature
- 24:29Indexing strategies
- 27:00Doc values for aggregations
- 28:23Time series frequency
- 29:45Arrays for measurements
- 31:09Joins for time and business
- 33:37User-defined functions for analytics
- 35:23When CrateDB fits
- 36:48Use cases where CrateDB is not ideal
- 37:32Time series challenges summary
- 38:37Q&A
related talks
36mCrateDB: A Search Engine or a Database? Both! #bbuzz
13mUnlocking Your AI Potential With Real-Time Data Unification
58mIntroducing ScaleDB, the New Clustered, Distributed Storage Engine for MySQL (Part 1)
18mBuilding Data Lake with MariaDB ColumnStore
25mData modeling, the secret sauce of building & managing a large scale data warehouse
23mWhen ms matter: Maximizing query performance in CrateDB
45mDealing with TBytes of Data in Realtime
30mCatching Waves with Time-Series Data