On Tue, Feb 24, 2009 at 5:52 AM, Jeff Hinrichs (JIRA) <j...@apache.org> wrote:
> PUTting json docs > 1MB causes Uncaught error in HTTP request:
> {exit,{body_too_large,content_length}}
> ------------------------------------------------------------------------------------------------------
>
> Key: COUCHDB-266
> URL: https://issues.apache.org/jira/browse/COUCHDB-266
> Project: CouchDB
> Issue Type: Bug
> Components: HTTP Interface
> Affects Versions: 0.9
> Environment: Apache CouchDB 0.9.0a747258
> Reporter: Jeff Hinrichs
>
>
> error displays itself when trying to PUT a json document that is > 1MB.
> First noticed in the python interface, confirmed with curl
>
> [Tue, 24 Feb 2009 13:30:00 GMT] [error] [<0.1113.0>] Uncaught error in HTTP
> request: {exit,{body_too_large,content_length}}
> 2
> 3 [Tue, 24 Feb 2009 13:30:00 GMT] [debug] [<0.1113.0>] Stacktrace:
> [{mochiweb_request,stream_body,5},
> 4 {mochiweb_request,recv_body,2},
> 5 {couch_httpd,json_body,1},
> 6 {couch_httpd_db,db_doc_req,3},
> 7 {couch_httpd_db,do_db_req,2},
> 8 {couch_httpd,handle_request,3},
> 9 {mochiweb_http,headers,4},
> 10 {proc_lib,init_p,5}]
>
> modifying src/mochiweb/mochiweb_request.erl > -define(MAX_RECV_BODY,
> (1024*1024))
> to something bigger, say -define(MAX_RECV_BODY, (1024*1024*16))
> alleviates the problem temporarily.
>
> issue confirmed by cmlenz on irc, he believed it to be a regression
This makes sense to me. It's probably due to my Mochiweb patch, which
created an alternate interface for streaming big docs. Using that
interface should fix things. I'll dig in here.
Chris
--
Chris Anderson
http://jchris.mfdz.com
