On Tue Jun 01, 2010 at 03:46:19PM +0200, Nils Breunese wrote: >Konrad Förstner wrote: >>Many thanks for your explanations, Randal and Nick. I am still >>astonished about the dimensions of disk space which this underlying >>data handling needs but I guess I have to live with it. > >One more thing: if you're going to manipulate lots of documents, see if >you can use the HTTP Bulk Document API: >http://wiki.apache.org/couchdb/HTTP_Bulk_Document_API > >For instance inserting lots of documents into a database will be much >faster and result in a smaller database when using the bulk API.
Thanks for the hint, Nils. The shown case was just a simulation. In the wild the updating steps are not taking place at once but scattered infrequently and the data migh be requested in between so it is not possible to use bulk action in this case. Cheers Konrad
