I have a software who does a heavy processing based on some files. I have to query some tables in SQL Server in the process and this is killing the DB and the application performance. (other applications use the same tables).
After optimizing queries and code, getting better results but not enough. After research I reached the solution: Caching some query results. My idea is cache one specific table (identified as the overhead) rows that the file being process need.
I was think in using AppCache Fabric (I'm on MS stack), made some tests it have a large memory usage for small objects ( appcache service have ~350MB of ram usage without objects). But I need to make some queries in these result table (like search for lastname, ssn, birthdate etc.)
My second option is MongoDb as a cache store. I've research about this and most of people I read recommend using memcached or Redis, but I'm using Windows servers and they're not supported officialy.
Using mongo as cache store in this case it is a good approach? Or AppFabric Caching + tag search is better?