An image service provides access to raster data through a web service. By default, an image service is always published with image service capabilities, and you can optionally choose WMS and WCS capabilities. Image servers can be federated with ArcGIS Enterprise deployments to enable imagery publishing and to perform distributed raster analysis based on the assigned server role. A federated image server can be beneficial when you need to perform distributed raster processing and analysis, and publish and share raster services. ArcGIS Image Server with an image hosting role indicates that the image server is dedicated to hosting the image services. To configure the image server with the image hosting role, it must have a registered raster data store. For details, see Register a raster store. The raster store registered on the image hosting ArcGIS Image Server stores the imagery data uploaded by the clients or generated by raster analysis tasks. The file share type of the raster store is required for the image hosting image server to be fully functional. Optionally, you can register an additional cloud store raster store.
Components needed for using a raster analysis server
ArcGIS Server provides raster analysis capabilities out of the box, and it boosts raster processing performance by distributing processing and publishing results to the server. Tasks that previously took days or weeks now may only take hours. Raster analysis requires ArcGIS Enterprise , which provides the infrastructure to organize and manage your distributed processing, storage, and sharing of raster and feature datasets, maps, and other geographic information on a variety of devices. The ideal deployment of raster analysis requires the set up and integration of three server sites to perform the primary roles of the hosting server, the raster analysis server , and the image hosting server.
The amount of data hosted by Internet servers and data centers is increasing at a remarkable pace requiring more capable and more efficient servers. However, physical efficiency does not necessarily correlate with computational efficiency. In fact, independent studies reveal that Internet servers are mostly over provisioned and still additional servers are deployed each year. Understanding the characteristics of the workload of servers is an essential step to efficiently manage them.