In adaptTo() 2014, I delivered a talk on “Integrating ElasticSearch with AEM”. It started with a brief introduction about ElasticSearch, its working and flexibilty when it comes to huge amount of data. Here is a brief review of topics covered after that:
- How data is analyzed and indexed on the server
- Integration approaches with AEM that included push and pull indexing
- Demonstration showing Free Text Search and Faceting with Geometrixx Site
- An introduction to some more features : Advanced faceting, Percolator, Geo facet.
- In the end there was a demonstration showing how easy it is to add/remove nodes in a cluster without having any downtime and any manual effort.
The presentation for the same is on slideshare.
As it was a 30 min talk, only some of the things could be included in there. Few points that worth mentioning about ElasticSearch that gives it edge over other frameworks are:
- ElasticSearch supports complex documents that could go any level deep as opposed to flat structure.
- Clustering is way easy and fast without using any additional plugin
- On the fly schema update : In development environment, you dont need to worry about the data and its type whereas when you are going on production you can always define schema to be strict so that data conforms to it. This saves lot of development time and effort.
- Percolator : Exactly reverse of traditional search. Here you have a query and your JSON document percolates through it. Much like “email labeling” that you see in gmail and other mail clients.
Hope you like it!!