Windows Azure - AppFabric Caching Service Introduction
In this article I would like to list the features of Caching Service of
Windows Azure AppFabric. As you know the services provided by
AppFabric are shown below. Why Caching is needed?
In the real business world scenarios the cost of building information
will be high. In this case if the same process is repeated for each
user requests then there exists some possibility for storing the
information for later reuse. This will increase the Speed of Processing
the request. In the server side this will free up the computing
resources to great extent.
Caching involves storing the information closer to the application logic. The data stored in the cache could be:
So caching is a good solution for the above scenario. But in the case
of distributed caching, if the application is load balanced and each
server needs to manage the update cache the problems become
exponential. Thanks to the AppFabric Caching as it handles all these
problems and the developer needs to concentrate on the application
Some of the core aspects of Caching are depicted below: Some Theory on AppFabric Caching Service Distributed Caching:
The Cache runs on N nodes with each node sharing the same configuration. The nodes will communicate each other. Cache Cluster:
Each Cluster consists of multiple cache server nodes. The first node
takes care of the cluster configuration and the next nodes can join the
existing cluster with no configuration. Cluster automatically adjusts
itself to the dynamic nodes addition or removal. Windows Service:
The caching is managed through Windows Service running on each node. Scalability:
Easier scalability of caching by adding nodes. More nodes results in more memory for cache objects Performance:
High Performance can be achieved in memory caching called Local Cache. Availability:
The existence of multiple nodes ensures higher availability on server failovers. Native API
.Net native code can be used to operation on the AppFabric Caching.
For ASP.NET through a provider change the session state can be cached
across a cluster. Provider Model:
Cache Configuration is based on Open Provider Model. The existing
provider models include Sql Server and XML file based network share. Projected Size:
Cache Size to be projected based on Small, Medium and Large sizes.
Small size include 1-5 nodes, Medium size include 6-15 nodes, Large size
include more than 15 nodes. Administration:
PowerShell can be used to administer Cache. The PowerShell can be invoked through code as well. Cache Events:
Caching supports events and the client can register for events which will be called back to them. Summary
In this article we have seen the AppFabric Caching service features.
In the next one we can experiment with code to access the Cache.