I have 300gb time series historical stock data, adding more each day. I build models and rely on sql to do lots of work. I have many bash scripts that run at points each day. I then use c++ to interface with the sql server and do things sql isn’t good at or fast enough at.
However my setup is flaky. I am running macOS on a Mac Pro with MySQL Server 8. It’s not reliable. Stops running. Won’t start up again. I have to move files around to convince it to work again. These same issues also happen on all the Mac hardware I have tried (Mac mini and a new MBP).
What is a more reliable setup that can use my Mac Pro hardware and the database living on an external USB drive? MySQL must be super stable on some platform? Should I switch to Postgres? I haven’t used it in 10 years but I assume they have a c++ connector.
Any advice appreciated.
Edit: It crashes because if cannot find a PID, even though it is there. It says it isn't the right one. I also was told that because the data was on an external drive this causes mysql to "lose itself". I was also told that perhaps there is a timing issue with mysql not finding the external drive in time and getting confused.
Edit 2: notated I have bash scripts running and I have tried other hardware with the same results.
But yeah, like akerl_ said, StackOverflow seems like the right place to ask otherwise.
[1]: https://www.timescale.com/