If _replicate was depreciated, there would be no way to schedule your own replications. Amongst other things, the _replicator database is inconvenient when a client machine uses different proxies on different connections.
> On 23 Sep 2013, at 5:59 pm, Benoit Chesneau <[email protected]> wrote: > >> On Mon, Sep 23, 2013 at 9:53 AM, Ido Ran <[email protected]> wrote: >> >> Hi, >> We are trying to write a ruby script that replicate a remote database to >> local CouchDB server. >> I thought that I should do an HTTP POST to the _replicate special URL and >> that I'll get back the replication information and the replication will run >> asynchronously but as we found out the HTTP POST request does not return >> until the replication finished (which can take up-to several minutes) or >> HTTP timeout occurs. >> >> Doesn't the replication request should be register in CouchDB and return >> HTTP response immediately ? > > If you want such thing you should better create a _replicator document. > > -benoit > > (note: we should really deprecate _replicate).
