incubator-couchdb-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexander Shorin (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (COUCHDB-1638) Memory overwhelming usage on attachment uploading when disk space ran out
Date Sun, 06 Jan 2013 21:56:12 GMT

    [ https://issues.apache.org/jira/browse/COUCHDB-1638?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13545510#comment-13545510
] 

Alexander Shorin commented on COUCHDB-1638:
-------------------------------------------

> 1.3 and 1.4 do not currently exist, so I've dropped them from 'affects'
I know and they was mentioned in dropdown menu as unreleased, but wouldn't this makes harder
to classify issue per releases?

Anyway, 
1.3.0: 88c52b23282ee37b1cc3720feee596fe1b938ea4
1.4.0: 4b20418391a9a77012ff74c8064f3da910c310fb
                
> Memory overwhelming usage on attachment uploading when disk space ran out
> -------------------------------------------------------------------------
>
>                 Key: COUCHDB-1638
>                 URL: https://issues.apache.org/jira/browse/COUCHDB-1638
>             Project: CouchDB
>          Issue Type: Bug
>         Environment: Linux, no OOMKiller
> CouchDB 1.3.0a-88c52b2-git
> CouchDB 1.4.0a-4b20418-git
>            Reporter: Alexander Shorin
>
> Uploading attachment that is bigger than disk space that left for CouchDB data causes
unstoppable memory consuming. For document updating in this case he just silently fails with
{{badmatch,{error,enospc}} in logs.
> Steps to reproduce:
> 1. Simulate small disk. I've done it with next commands:
> mkdir /var/lib/couchdb/sandbox
> chown couchdb /var/lib/couchdb/sandbox
> mount -t tmpfs -o size=20m tmpfs /var/lib/couchdb/sandbox
> 2. Create database with name e.g. sandbox/deadman
> 3. Create new document and upload attachment with size >=20MB
> 4. watch free and get ready to kill CouchDB
> As for me CouchDB may easily consume 8GB RAM and 3GB of swap for described above conditions.
> Expected behaviour: not so fatal.

--
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

Mime
View raw message