[jira] Commented: (COUCHDB-690) replication fail -- couchdb crashed

["Daniel Bechler (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/COUCHDB-690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12863769#action_12863769 ] 

Daniel Bechler commented on COUCHDB-690:
----------------------------------------

I'm having the same issue (and same error log). This happens frequently on a small (<10k docs), but write-heavy (very high update frequency of all docs) database, when replicating from a server in the USA to Europe. Running CouchDB 0.11.0 on both servers.

> replication fail -- couchdb crashed
> -----------------------------------
>
>                 Key: COUCHDB-690
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-690
>             Project: CouchDB
>          Issue Type: Bug
>          Components: Replication
>    Affects Versions: 0.10.1
>         Environment: linux  2.6.30.7 - debian 5.0
>            Reporter: linkfluence
>            Priority: Critical
>         Attachments: couch.log
>
>
> We have a database on host A with 8.5 millions document. The size of the database is ~450GO. 
> We first tried to start a continuous replication on a second host B. The replication stoped after only 1Go have been copied, and the replication never started again.
> We then copied the database file from host A on host B. When the file was copied, we started a replication from A to B, then the couchdb on host B crashed. It tooks a long time to fetch a list of IDs, then it appears in the logfile that a  time out occured on host B, and immediatly after the couchdb instance on host B crashed.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.