3

I am pretty new with Druid and I don't get my answers regarding the comparison with ElasticSearch. I found this link: druid vs Elasticsearch but it does not give the differences and advantages.

Can anyone explain me that or give me some links that I didn't find on google ?

Thanks in advance.

J

Simeon Leyzerzon
  • 18,658
  • 9
  • 54
  • 82
  • It's apples and pears. Druid is for aggregating data, Elasticsearch is a search engine. It's not a useful comparison. Elasticsearch could be compared to e.g.: Solr, Sphinx or barebones Lucene. Druid is pretty much alone in the aggregation department afaik. Let us know what you want to accomplish and I'm sure you get a more satisfying answer. – Geert-Jan Oct 27 '16 at 14:35
  • Thanks for explanation. I am trying to build a lambda architecture where I have to select a data base for my serving layer. Indeed, on this serving layer I want to request in real time a large amount of data in order to filter, do some aggregation and other computations. I selected ElasticSearch instead of Cassandra and MongoDB but some of my collegues indicated me Druid. However, ElasticSearch also does aggregation. Thus, is Druid better than ElasticSearch for aggregation and filters ? –  Oct 27 '16 at 14:58
  • 1
    if your colleagues will help you with the support of the Druid, it is better to choose for Druid, but I advise you to try it out first. At druid too high entry threshold. "Druid is extremely difficult, complex and time-consuming to initially set up, and operate in production. So unless you really need to ingest 1M+ events/sec AND you have a dedicated team to take care of it, just use Elasticsearch" https://github.com/Banno/druid-docker/issues/3#issuecomment-190749607 – henb Oct 31 '16 at 19:58

1 Answers1

1

Druid commiters have a nice short text - with some bias, of course - about this. Maybe will help you.

In summary, Druid is great with timeseries data because it have Roll-up funcionality and is based on Historical Nodes.

ElasticSearch is a full-text search database on top of Apache Lucene. Have features to Recomendations and Data Mining like More Like This and Fuzzy query.

Allan Sene
  • 459
  • 4
  • 15