A nice intro video on ElasticSearch from Air Mozilla. It is ~56min.
—Jason
Great write-up on the Elasticsearch course.
- Stop using Query_String
- Prefer the Bool filter over And/Or/Not filters (usually)
- ElasticSearch + SSD = match made in heaven
- Design filters that are cacheable
- Scale out, not up
—Jason
I learned a lesson today while developing a new plugin (a River plugin) for ElasticSearch. It seems that when loading/running plugins ES catches any uncaught exceptions AND doesn’t log them. This made it really hard to debug an issue I had where I didn’t have the right JARs included in the plugin assembly. Here is an example assembly.xml file.
The problem would only manifest itself when I started reading data from a remote service. Inserting lots of log statements helped me identifiy where the code was silently failing, and then I used Eclipse’s remote debugging tool to see what the issues were. It turned out to be ClassNotFound exceptions.
If I get approval I will give more specifics on what the River does and possibly open source it on GH.
—Jason
This conference looks really interesting and is probably worth keeping an eye on.
Lucene/Solr Revolution 2013 The Westin San Diego April 29 - May 2, 2013
—Jason
Short article on using the Apache Blur REPL for interacting with it.
If you haven’t checked out Blur before, you definitely should. It is an open source search engine built over Lucene and Hadoop. It provides a Thrift API for queries, and it mainly uses MapReduce for ETL/ingest.
Here are some of the features listed on its website:
- Fast data ingestion
- Hierarchical data storage
- Record-level access control
- Paged results
- Quick search
- Boolean search logic
- Fuzzy searches
- Wildcard searches
- Facets
- Term statistics
- Term lists
—Jason
Chef Recipes for Installing Elasticsearch on Linux servers
Great article on Lucene’s new-ish (3.1) feature for supporting Near Real-time searches.
—Jason
Brief article about an upcoming Lucene feature. The LiveFieldValues class…
—Jason
