Before actually (re)writing sofware to be used in the Microsoft Service Fabric I need to get to know the basics about a few things. Today I was wondering how to actually persist data. In the ‘old ages’ (yesterday..) I was always using some kind of database to store application data. This is in a microservices architecture not a bad thing, but you need to think about the data architecture before doing so.
First of all I was wondering if I really needed to persist the data for the application I had in mind. (Sorry, I can’t go in to details about the application itself) The data stored in the application is mainly configuration data and some things needed for reporting and logging. The configuration is not much, but needs to be persisted and usable even after a crash of the entire cluster. The logging/reporting data is quite extensive, but is volatile. It needs to be available for some time (1-3 months) but if lost no harm is done.