 |
Whether you are deploying to a small department or looking to configure a company-wide installation, Coveo Enterprise Search is capable of meeting the most demanding requirements. The solution can manage up to 50 million documents per server and returns high quality results in less than a second. Requiring less hardware for a simpler configuration, Coveo Enterprise Search offers better functionalities and a total cost of ownership that is significantly lower than the competition. |
|
Query performance and High Availability
Coveo’s unique use of mirrors and network load balancing (NLB) guarantees the high availability of documents. |
| Through our mirroring system, if one mirror fails, others are in place to continue serving queries. By doubling the number of mirrors placed under NLB, the number of queries that can be answered per second doubles. As indexing progresses, mirrors are automatically synchronized and updated. Multiple front-end web servers can be used for sites facing heavy traffic. An index footprint is typically less than 30% of the original document size. |
 |
|
Distributed indexing using Index Slices
The indexing process is distributed in many index 'slices', where each 'slice' indexes part of the content. These 'slices' can be hosted locally (on local or network drives) or on separate servers (through IP connection). |
 |
|
Federated Search & Geographically Distributed Indexing
Coveo Enterprise Search can federate queries from multiple remote search servers and merge the results into a single page while preserving all the relevancy information required to rank results coming from different sources. |
 |
Configuration 1 - Federated search with a remote location
- In this configuration, there is no copy of the index between remote locations, thus reducing the network load
- When a query is executed in one location, it is executed on remote indexes with the results federated on the local server
Configuration 2 - A remote mirror index using geographically distributed indexing (GDI)
- In this configuration, a copy of the remote index is kept on the local server
- Queries are executed using the local index and copied on the remote index
|
|