What is the caching mechanism in Live Score context?
book
Article ID: KB0077477
calendar_today
Updated On:
Products
Versions
Spotfire Statistica
13.3 and above
Description
Caching in Live Score is mainly introduced for performance reasons. It takes a relatively long time to instantiate and fully load a workspace for the first time it is called. When a workspace is "cached" (i.e. an instance of that workspace is retained in Statistica worker process), which happens on its first run, it then could be matched and quickly retrieved on subsequent requests to speed up the processing time of running the same workspace. When the cache is cleared, the workspace starts from a clean state.
Resolution
An example of this Cache line in the WSDL script :C:\WebSTATISTICAPub\RepositoryRoot\System\Scripts on the Live Score Server. Note: The below SOAP example is specific to calling Live.svb from the SOAP request. ======= <ClearCache>1</ClearCache> ======= The <ClearCache> tag can be set with value 1 or 0. This flag is not "hard-coded" into the system but is just a condition implemented in the sample Live.svb server-side Live Score script to trigger cache clean-up, in which case the workspace model has to be loaded again. When ClearCache is set as 1, the workspace will run from a clean state. When ClearCache is set as 0, filing a SOAP request will first search if the workspace has been loaded before, and if yes, it loads from the cached memory and speed up the request.
Since the models are usually retrieved from Enterprise, the cached Enterprise copies might go out of sync if the saved version of the Enterprise model is updated. In that instance, please set the ClearCache as 1 to load the workspace again for a fresh start.
Issue/Introduction
This article explains the caching mechanism in Live Score context.