[
https://issues.apache.org/jira/browse/COUCHDB-1777?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13642904#comment-13642904
]
Adam Kocoloski commented on COUCHDB-1777:
-----------------------------------------
Cool. That's a different (and significantly more complex) code path. The most
likely scenario is that CouchDB/mochiweb believes it has not received all bytes
on the connection at the time it issues the 409 Conflict response, in which
case it will definitely close the connection.
I agree that adding a Connection: close response header in that case is a
polite thing to do.
> 409 response vs persitent connection
> ------------------------------------
>
> Key: COUCHDB-1777
> URL: https://issues.apache.org/jira/browse/COUCHDB-1777
> Project: CouchDB
> Issue Type: Bug
> Components: Database Core
> Reporter: Marek Kowalski
>
> CouchDB closes the persistent client connection after responding with 409
> code. This behavior is not necessarily against the RFC 2616, although it
> makes it more complicated to use pipelining.
> If its really necessary to close the connection, it would be nice to at least
> put the "Connection: close" in the response headers. If the connection can be
> closed without informing client that this is going to happen, the client
> might start writing the new request to the connection. Again, according RFC
> 2616, this situation should be handled by the client, although the header
> could be added "as a courtesy".
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
