24 ms·
Ask HN: Dynamic data in a database
I am trying to figure out whether or not I should make the switch from a standard database to something new and fancy like mongodb, but I am still unsure how my problem fits into either solution.
Essentially what I have is data that grows with time, so a column based database seems likely to cause headaches in the near future. I should also note that I receive all my data in a JSON format, so I was leaning towards a mongo solution given those facts. It seems pretty keen on storing data in that object based format.
I think my only concerns are how do you do match up data from one collection to another. Say for instance I wanted to organize my data in such a form so that I had an entry collection, a hash data collection and then a results collection. How would I associate those details with a given entry within the entry collection? Or would I just store it all together?
- Skywing 16y agoYou might also want to check out Redis. I use both MongoDB and Redis. I don't quite understand your requirements or data schema, though.
- maxdemarzi 16y agoBrandon, I think you need to give us a bit more information about your data in order to get meaningful answers. When you say "data grows with time", you mean the number and type of fields for an "entity" change right? Mongo or any document db fits that purpose well. I am not sure I understand your question about matching things up? You can store all "related" objects together in one document, you can "link" objects together like done in Riak (or with a list of incoming and outgoing relations in the document) or you can use a graph DB (like Neo4j).