I have a setup using Dapper where I have many databases that are structured identically to each other. There could be potentially thousands of them, and the way I have to connect to them is by using a different connection string on each.
I know that Dapper caches some information for each query related to reflection and maybe other stuff, but I was wondering if having many different connection strings will result in a notable increase in memory usage over a single DB, or if the cache is more for reflection data and not connection-specific.
Thanks in advance.
Dapper came out of some of the Stack Overflow development, and what you describe is exactly how Stack Overflow (or rather: the Stack Exchange Network) works - i.e. a multi-tenancy process that talks to any of hundreds or thousands of connection strings depending on the specific request. So: Dapper deliberately doesn't do anything that is connection or connection-string specific; all the caching is based on the look of the type, command or results.