I am starting to develop an MVP, and since %99 of MVPs are destined to fail, I have decided NOT to spend a lot of time thinking of the db layer, at least initially.
I already made a few MVPs in the past, and every time I had to spend a lot of energy and time on DB issues, only to have the MVP (and all the work that came with it) fail .
So I am thinking of starting with a RAM only db and only if the MVP reaches more than X users, I will actually start thinking about a real, physically independent, secure and reliable db .
I can 'get away' with a RAM-only db by creating a json txt file db that will be loaded upon runtime and will be backup-up on an interval basis (eg: every 1 min the RAM db will be saved into the json txt file).
To summarize:
pros: Extremely fast development of an MVP
cons: Not sustainable, not scalable at all, only fits for the initial stage of the startup .
Any thoughts ? experiences ? life lessons ?
ANY info will be appreciated .
0 comments
—
Ask HN: Building an MVP with a RAM-only database – a good idea? · HackerTrans
So I am thinking of starting with a RAM only db and only if the MVP reaches more than X users, I will actually start thinking about a real, physically independent, secure and reliable db .
I can 'get away' with a RAM-only db by creating a json txt file db that will be loaded upon runtime and will be backup-up on an interval basis (eg: every 1 min the RAM db will be saved into the json txt file).
To summarize: pros: Extremely fast development of an MVP cons: Not sustainable, not scalable at all, only fits for the initial stage of the startup .
Any thoughts ? experiences ? life lessons ?
ANY info will be appreciated .