I looked into Apache Druid for my company's event aggregation and serving. We decided to go with Apache Pinot instead. A major reason was serving latency. Most of the docs I saw that talked about Druid latency (and their goal for latency) made it seem like 500ms was okay for query latency speeds. Pinot seemed more focused on query latency. 50ms was more acceptable for my team's latency goal. I ran some basic latency tests and Pinot was faster (2mil and 22mil events, no real optimizations). I think Pinot's star-tree index would perform great for my team's use case.
Druid's intro flow was awesome. The tutorials worked very well. I hit hiccups with Pinot's intro flow. The Pinot team has been very responsive on Slack to issues.
Similar to @quietgolfer, we also have strict requirement on serving latency, and we benefited from Pinot's star-tree index. We had a table with more than 1 billion events, and we had to do DISTINCTCOUNT and SUM with GROUPBY. Initially without optimization, queries timed out (after 15s). After star-tree index was added, the query returned in 500ms.
Based on our earlier benchmark, Pinot query latency without star-tree is already slightly better than Druid: https://image.slidesharecdn.com/lnkdmeetupslides0426-1705021.... Though that benchmark was done years ago, the result seems consistent with the recent test run by @quietgolfer.
Druid's intro flow was awesome. The tutorials worked very well. I hit hiccups with Pinot's intro flow. The Pinot team has been very responsive on Slack to issues.
Here are some notes:
https://docs.google.com/document/d/1GiB8zoiJ4Qs10A1LdIr6lJGW...