DBSight is super easy to use and high-performance full-text search platform. Users usually only specify SQLs and choose rendering templates. DBSight takes care of the rest to run the search service. DBSight is based on Lucene library, with faceting, sorting, ranking, caching built-in. It runs much much faster than database search, and outperforms or is on-par with any other Lucene solution. DBSight is not only great for high-traffic web sites, but also ideal for Enterprise with multiple data sources, changing requirements, access control, worry-free deployments, etc.
Only SQL knowledge is required to create a search! DBSight would generate schemas based on your SQL, and you do not need to write any Java or XML. Even incremental updates will be managed by DBSight. SQL extractor engine is efficient due to cache and batch selection.
The template engine is fully customizable with powerful scaffolding system. Pre-built solution includes Search Engine, Data Grid, JSON/JSONP, Excel, CSV, XML, etc. You can continue to adjust the css, layout, or add/remove features, to meet your taste.
Only SQL knowledge is required to create a search! DBSight would generate schemas based on your SQL, and you do not need to write any Java or XML. Even incremental updates will be managed by DBSight. SQL extractor engine is fast with caching, batch selection features.
Several templates to choose from, ranging from basic XML, CSV, JSON, JSONP formats to templates full of bells and whistles, with features including Suggest As You Type, Infinite Scroll, Data Grid, Facets, Pagination, Sorting, Highlighting, Summarizing, etc.
Simultaneously search multiple local or remote indexes! Aggregating data from on multiple sources in a single page!
Multiple ways to count results by each category, or by customizable value ranges, or by date on year/month. Facet can even have hierarchical relations, e.g., facet on state level first, before facet on city level. Facet can also get minimum, maximum values, or even sum/average based on a separate field.
Based on high performance Lucene engine, DBSight add data compression, auto warming, more caching, to ensure performance even with more features.
If your data grows rapidly, you can split large index into several indexes on local or remote machines, and search them together as if it is one index, but with improved performance.
Automatically provide suggestions as you type, based on existing actual data.
UI driven friendly design, most features are enabled/disabled by mouse clicks.
Adjust search results ranking by its own value, by its freshness, or by the matching field. All done via UI.
The search front-end, indexing processes, and search processes are run in their own Java processes. Common memory problems, e.g. long GC time due to memory fragmentation, are gone. One slowness would not affect the whole system.
Creating one new index is so trivial and easy ...
Automatically filter out results that users don not have permission to access, based on users' roles.
Customize the default scaffolding by adding your own logo, css, layout, etc. Ensure consistent look and feel for your own department or your own consulting services.
Pre-defined data sources allows easier deployment to production, and easier to manage database accesses.
Moving the configuration is as easy as 2 steps: export, then upload.
Your developers may want to write their own Analyzer, Similarity, etc. DBSight makes it easy.
Merging takes CPU and disk IOs. For many sites, you can configure DBSight only does these work on non-peak hours.
If you want to rank new results higher, it is just one click away!
To upgrade, just install the new versions again! No customization would be lost!
Search API is based on Google's Protocol Buffer. You can use any language to integrate. Java library is provided by default.
Content can be submitted via HTTP post and instantly searchable!
Super efficiently get top queries in real time! Generating search usage histogram for every hours!
Analyzer can be customized, tested, compared just by UI.
DBSight can be configured as stand-alone, or sharded to multiple servers. It can also be configured as 1-indexer-multiple-searcher, or multiple-indexer-multiple-searcher, or multiple heterogeneous instances. Search can run across multiple different instances efficiently and smoothly.
When new index is ready, a notification will be broadcasted. Subscribers will pull the new index over. Also, you can schedule the replication if the indexer and searchers are on different network.