Speed has and will always be a critical factor in developing fast and responsive web applications. The cost of computing power has seen a rapid decrease over the years allowing for fast and powerful pc’s capable of handling much of the server side processing, thus allowing most web apps to load faster and respond quicker to client requests. But then came big data. In the web mapping world where huge amounts of data needs to be served to the client, maintaining a high level of performance and usability can only be possible through pre-computing on the server.
A well known example of pre-computing is Map Caching in which tiles are pre-generated and cached on the server allowing for faster map services.
Over at the ArcGIS Server blog is a nice post highlighting the importance of pre-computing in a GIS server environment and techniques such as
- Eliminating on the fly projection
- Pre-calculating geoprocessing results
- Building spatial and attribute indexes
- Converting labels and annotations
that can be employed in publishing faster GIS web services. These practices are not just limited to the ArcGIS product domain but can also be applied in any web mapping environment.
Boosting performance through pre-computing
