How does CouchDB perform for a regularly updated dataset?

Question

I plan to use CouchDB for the project. But since the query mechanism includes recording representations (which are very similar to indexes on regular RDMBMS), I was wondering if the document database could be constantly updated (writing a heavy database), would CouchDB work well compared to a regular DBMS? Or do we sometimes have to compactly / re-index the system to make it work faster?

Answer 1

You might think about the pros / cons of the CouchDB presentation model this way. (CouchDB hackers may disagree, but IMO is accurate enough for users.)

• The view function always performs a full “table scan” when it is first created (like BDBMS BTW).
• As long as they have no side effects, the display and reduction functions can be arbitrarily complex.
• Each document and map / reduction result is cached and never calculated again
• If you add or modify a document, it (and only it) will be recalculated (and cached) for this view

Given this, you can draw some conclusions about the performance of CouchDB:

• There is never a re-index phase for an entire dataset, just incremental to update a document
• Changing the view function forces the entire index to be recreated.
• CouchDB RDBMS , , /.

, YMMV , " ". .

• , RDBMS , . , , , .
• , , , CouchDB (, , Hadoop) , .
• , NoSQL . , CouchDB-Lounge . Hadoop , : , , CouchDB - -, DIY; Hadoop , ..

, !