Search

Cyberitas Search Technology

A search is a search, right? Wrong. A slow search loses customers. Surveys continually show you have 4 seconds or less to capture or keep someone's attention on the web. Most web sites rely on database search technology, and high overhead logic that limits their search potential. Have you been to a web site that said "Please wait while we perform that search..."? It wasn't a site using the Cyberitas search technology, which enables true real-time searches and data lookup at speeds simply not possible with other strategies. Utilizing associated data links as well as optimized access algorithms, the Cyberitas search technology can complete searches faster than many platforms can even start searching.

Fast Search quickly, with intelligent cross-data links, highly optimized data structures, and no database connection overhead. Most searches take only milliseconds, even in single server installations.

Reliable - All search technology servers have been refined by Cyberitas over the last 10 years, and are stable self monitoring processes. Internal controls guarantee data integrity, and up-time. Because any single cluster node can be replaced or restarted independently, maintenance can be performed undetectable to end users -- even in high load conditions.

Scalable - Cluster capable, the Cyberitas Search Technology is designed to run at the edge of your cluster, and have a linear speed increase as servers are added to the cluster. Unlike monolithic database solutions, there is no penalty to adding more nodes to your clusters.

Rich Data Storage Algorithms - The Cyberitas Search Technology stores data in several highly optimized structures for extremely fast access. To demonstrate the power of these algorithms, the small example on the right features nine names mapped in one of the same ways that they are stored in the search technology's shared memory. By following the "lines" to read, instead of reading sequentially, you can find out if any name is in the list by reading no more than N*3 the number of characters in the name -- even if there were many millions of names in the list! You will also see it takes only 20 characters to store all 9 names, which would take up 42 characters if stored individually.

Once you reach a desired result, all associated records are immediately available without searching by following the intelligent cross-links created by the Cyberitas Search Technology. In this example, each of the nine names is linked to a city contained in another data set. If you searched for a name, you would immediately have access to the city record, and all data associated with or linked from the city record.

In addition to searching faster, the Cyberitas search technology protects your data -- even when a database goes down or needs maintenance. Each machine at the edge of your cluster can run searches autonomously. This allows you to perform maintenance on a cluster node at any time, even if it is the database server, without affecting end-user searches. Once the database is back up, the search technology connects, and refreshes its data store.