title | date | draft | tags | summary | ||
---|---|---|---|---|---|---|
today we are going to share a new paper about druid
|
2019-12-11 |
false |
|
Hi all, today we are going to share a new paper about druid. http://static.druid.io/docs/druid.pdf It's related to the previous share about scuba. They are all column-oriented database storage. Like scuba, The Druid database also has an in-memory data storage strategy.
|
Hi all today we are going to share a new paper about druid.
http://static.druid.io/docs/druid.pdf
It's related to the previous share about scuba. They are all column-oriented database storage. Like scuba, The Druid database also has an in-memory data storage strategy.
Highlights:
- Database module roles. (include Real-time node, Historical Nodes, Broker Nodes, Coordinator Nodes)
- A complex but very useful time window design for storage "segment" in memory.
- read data from Kafka and can be query in multiple hundred milliseconds after event happen.