In simple terms, Cache is a temporary storage of the data, from where the user can re-fetch it more than once, without the need to re-process the request. One example of cache is your temporary internet files in your local folder, which allows you to pick pre-existing pages without going to the world-wide-web.
Layers of Caching
RAM Caching
You may also get a view of this in the latest trend of Memory-base analytics, where a whole OLAP cube can be maintained within RAM. Even without in-memory analytics, the systems have intelligence memory handling capabilities, whereby the data once pulled into RAM, is maintained there till it is over-written by the new data.
Data base Caching
This caching is of the datasets generated by the database level queries.
Application Server Caching
The application server though which you access your OLAP server, is a layer sitting between the user and OLAP server. For faster delivery, this layer can also contain the data-sets OR pages.
Client Level Caching
This caching is maintained at client desk-top, which obviates the need to use the net-work traffic.
Types of Cache
- Calculations- These are aggregates calculated on the fly and stored
- Data - This is the data in terms of data-base tables
- Data File - This is more in terms of flat data files
- Indexes - If both index as well as the data is cached, your ability to search data within cache is enhanced.
-
Pages: These are the complete pages (like web-pages).
A smart caching system in an OLAP would intelligently decide OR provide suggestions on:
-
The amount of memory (RAM- which is generally automatic) OR disk space to be allocated for cache at different layers.
- Which data to be cached in which layer: This is done with the help of constantly checking up the versions of the data.
- Which type (OR a combination) of cache to be created in which layer.
|
