I need to create a specialized caching service (either a WCF service, or a remote .Net service, or another implementation) for caching data (monitoring data from a .net3.5 time delay measurement application) from a database (SQL Server 2005), which can then be used 2-3 Windows services (.net 3.5 WCf services), so they don’t need to re-access the database. Thus, basically this is an additional level of service level between these services and the database to reduce performance when calling the database. The data in the Database is large and very dynamic (events that are constantly inserted into the database throughout the day) that we want to cache. Each of the services consuming windows depends on another (one service extracts data from one table in Db and aggregates the data and puts another table in the database,which will then be consumed by another service). We have the following requirements:
- It can have either a Push or Pull model to push or receive data to / from the database and services.
- The cache should reduce the overall impact on the database and delta data loading, and not recharge the data completely
- Cache data should be updated if the original data in the database has changed. Sql cache expiration policies must be defined.
- Must have asynchronous method calls to retrieve data from the database and update the cache to minimize latency.
- The cache must support concurrent queries to a single database
- CPU and memory usage should be maintained at optimal levels so that it does not adversely affect other services.
We do not have a clustered or distributed environment and should not be a very scalable solution.
, , , :
.
,
KK