Understanding IMPAX caches

When images are sent into IMPAX from an acquisition station or retrieved from an archive, they must be stored temporarily in an accessible location. This temporary storage area is called a cache. A cache is a set of directories on local or external hard disks.

IMPAX supports two types of caches: image and web. The image cache contains what in the IMPAX Client is called original images—the images as they are sent from the acquisition station. The web cache contains copies of the original images that have been compressed with wavelet encoding by the Curator component. In many cases, the version of image in web cache is faster for Clients to load.

You use the Cache Manager in Administration Tools to create cache volumes on stations and to set the priority for cache usage. The archive must have at least one cache it can use. If using the Curator component, it requires at least one web cache.

Prior to IMPAX 6.5, the cache structure was flat (each cache volume contains one directory) which limited the cache size, because once a certain number of objects are in the directory, access to the cache can become very slow. Large sites may resolve this by deploying numerous cache volumes, which can be difficult to manage.

As of IMPAX 6.5, a hierarchical cache structure is supported for image and web caches, permitting larger cache volumes. The old flat cache structure continues to be supported; only new images arriving in the system or existing images retrieved from archive are written to cache using the hierarchical structure. However, the cache migration tool, which is included in the standard IMPAX install packages, allows a site to migrate its existing caches if it would like to immediately take advantage of the hierarchical structure.