incubator-couchdb-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Emmanuel Decitre <deci...@gmail.com>
Subject Re: bypass mod_couch
Date Tue, 07 Dec 2010 21:57:50 GMT
Zach,

Alright. Another good reason to learn erlang.
And maybe also EPAPI to interface my python VM to hovercraft though a C
extension...

E

On Tue, Dec 7, 2010 at 10:32 PM, Sebastian Cohnen <
sebastiancohnen@googlemail.com> wrote:

> Emmanuel,
>
> you can bypass the HTTP stack with hovercraft. but you have to know some
> erlang for this though.
>
>
> [1] https://github.com/jchris/hovercraft - not sure if this is working
> with current couchdb versions.
>
> On 07.12.2010, at 22:14, Emmanuel Decitre wrote:
>
> > Hi Zach,
> >
> > Thanks for the info.
> > PUTing bulks will definitely speed-up the whole.
> > But since I have around 1E9 documents, it might
> > still be worth bypassing the networking layer.
> > I'll however give the bulk PUT a try.
> >
> > Cheers,
> > Emmanuel
> >
> > On Tue, Dec 7, 2010 at 9:45 PM, Zachary Zolton <zachary.zolton@gmail.com
> >wrote:
> >
> >> Emmanuel,
> >>
> >> Are you PUTing each document into the database individually?
> >>
> >> If so, you should try the _bulk_docs API:
> >> http://is.gd/imdGF
> >>
> >>
> >> Cheers,
> >>
> >> Zach
> >>
> >>
> >> On Tue, Dec 7, 2010 at 2:34 PM, Emmanuel Decitre <decitre@gmail.com>
> >> wrote:
> >>> Hi there,
> >>>
> >>> I am a newbie in the couchdb community.
> >>>
> >>> I have a huge set of gis data I am storing using the document REST API.
> >>> But it takes hours. I was just thinking whether it is possible to
> >> directly
> >>> access the storage engine (make_doc in couch_db.erl ?) to feed the
> >>> beast offline before starting the http engine...
> >>>
> >>> Any hint ?
> >>>
> >>> Emmanuel
> >>>
> >>
> >
> >
> >
>
>

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