Posts

Showing posts with the label performance

Split Database for Better Performance

Image
Gladinet Cloud Enterprise (GCE) is an Enterprise File Sync, Share, and Collaboration solution.  During POC (Proof of Concept) stage and early production stage, it probably doesn’t matter how the configuration information and runtime data is stored. However, as more and more users are getting onto the system. We need to start thinking about split the database for better performance and adding worker nodes for load balance. Two Types of Traffic GCE database is one of the most critical components of the GCE platform as it contains two distinct type of data. The first type is static which includes for example username, email, and the user’s published shared files and folders. This type of data is relatively small and can be estimated at 5000 bytes per user. Therefore, for a 2000 user implementation, the database size for this static data can be estimated at 10 MB. The second type of data is volatile for example; file change log, audit trace, and the file list. The database size fo...