-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
I have considered using 2 raspberry pies. One would just handel rebooting the other. I am also looking at fog backup servers, but I think that might incur too much data usage on my cell modems.
If you want to really get paranoid, then you can write a monitoring app that uses machine learning to do the log analysis and detect anomalies in your system. There are some open source tools available, like this for example. Also you can train the network for your specific use case and then just have the service running the inference on your logs and a pre-trainer model that is running on system logs. Then you really get in paranoid mode.
That depends on your application. Personally, for custom things that I do which are not products, I'm using nanopi SBCs. I'm also the maintainer of the yocto BSP layer for those boards. Using yocto I can have a full control on the distro, but it's more difficult workflow compared to use any ready-to-go distro for your SBC.