this may not be the proper place to ask this question, but I was not welcome posting it on StackOverflow so I thought this may be a good place. If this is the wrong place, please refer me to the right place. Anyways!
Right now I am using a distributed database to store encryption keys. We have multiple nodes that all need to use these whenever a device connects, so all nodes must be able to quickly access and retrieve the keys. This means that if the database is down, our service doesn't work, so the uptime is incredibly important. Reads happen 50x more often than writes, so consistency is not super important, as the keys don't change often. The most amount of data storage we could possibly need is 8GB and that would cover 10m users - so the data per user is incredibly small.
I realize now, after a long while of using this distributed database, that it may not be the right solution for my needs. Because our data overhead is so low, keys might as well be stored in RAM and a copy on disk. Also, distributed databases are usually built to store LOTS of data, so they aren't necessarily geared towards our idea of small data + quick access times.
So, I guess my question is, is a distributed database necessary here, or would a relational be better for us? Maybe an NDBM or relational?