incubator-couchdb-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Der Engel <>
Subject Re: Would this be a sane use case for couchdb?
Date Thu, 10 Apr 2014 00:45:36 GMT
Thanks for you answer.

When I said I won't need a cluster I meant from the "scaling" side of
things, but will probably need a master-replication setup for backups. Is
it better to do analytics on the backup?

So taking your address card example, suppose you have a web app  where
multiple users can log in and each one can created their own list of
address cards, if understand you correctly, each user should have only one
document for their list of address cards? or one document per address card?
What about if in the future, users want to add more information about their
contacts? like their favorite books (with price, author, title) for each
contact, should all of these still be in the same document?

If that is the way to design/model data for CouchDB then how big is too big
for a document? For example, in my case, invoicing software, users are
going to create hundreds invoices every year, so keeping all the invoices
on a single document is going to make the document big overtime, is that

I'm a newbie that has never dable with any database, so I'm probably asking
the wrong questions here, is there some design principles/examples
somewhere (books are ok) for modeling data for CouchDB? or maybe there is
some generic modeling that applies to CouchDB too.


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message