Search engines

This family consists of NoSQL data stores crafted for one specific goal, the indexing and search into their data content using complex, full-text and / or distributed search queries.


ElasticSearch

Positive

Negative

  1. Easy scale-out
  2. Sharding and index replication
  3. Multi-tenant
  4. REST API
  5. Support complex / nested documents
  6. Full text or keyword search, faceted
  7. Self-contained cluster
  1. No security without (commercial) add-on
  2. Its distributed nature can have negative effects on consistency
  3. Own query language
  4. Once index is created, number of shards cannot be changed
  5. Not a main datastore, but for information like metadata to be searched.
  6. Support only JSON

Recommended use:
  • To build (web) applications with a rich search interface
  • Real-time centralized search and correlation of systems and application log messages
Continue to explore this site to know more about Elastic Search.


Apache SolR

Positive

Negative

  1. Highly scalable and fault-tolerant
  2. Sharding and index replication
  3. No re-indexing after adding capacity
  4. Multi-tenant
  5. REST API
  6. Optimized for high volume traffic
  7. Full text or keyword search, faceted
  8. Bundled web admin interface
  9. Support JSON, XML, CSV
  10. Inter-index joins
  1. One set of fields per schema, one schema per core
  2. Cluster depend on separate ZooKeeper server
  3. No automatic shards rebalancing

Recommended use:
  • When building an Enterprise Search Engine