It looks like this must be the successor to Deliroll, which was described in an absolutely <i>fantastic</i> talk from 2013[0]. I've been hoping this would be open sourced for years. Really glad to see this!<p>[0] <a href="https://www.youtube.com/watch?v=rXj5nayS7Yg" rel="nofollow">https://www.youtube.com/watch?v=rXj5nayS7Yg</a>
I was just watching this[0] excellent talk from one of the Heap Analytics engineers on their use of CitusDB + Postgres to do event processing.<p>What use-case would TrailDB be the obvious, hands-down way to go (vs maybe Postgres + Citus)?<p>[0]: <a href="https://www.youtube.com/watch?v=NVl9_6J1G60" rel="nofollow">https://www.youtube.com/watch?v=NVl9_6J1G60</a>
I'm so glad AdRoll finally open-sourced this. TrailDB is by far the easiest way to process trillions of events on a single machine. Can't wait for them to start publishing their internal ecosystem around it too.
How does this compare to PipelineDB? I'm pretty sure the cofounders of Pipeline are ex-AdRollers, so there should be a lot of overlap since it was made with the same problem in mind.
I'm assuming this is primarily used for analytics ? Metamarkets another ad company open sourced Druid for a similar purpose. I'm curious as to the differences in these solutions.
Might have to check these out. I'd be interested the match of storing data like this.<p>Also, another event DB project: <a href="https://github.com/benbjohnson/skydb.io/blob/master/source/blog/2012-04-30-introduction-to-behavioral-databases.html.markdown" rel="nofollow">https://github.com/benbjohnson/skydb.io/blob/master/source/b...</a>
Very interesting. I wonder how this compares to writing events to a Parquet (<a href="http://parquet.io/" rel="nofollow">http://parquet.io/</a>) or Avro (or both) file - this way all manners of distributed tools like Hive or Spark could already read it.
At which point in the curve of users-in-your-app it becomes better to use aggregate queries like this instead of just looking at each user individually?