On Friday 27 June 2014 05:37 PM, Nux! wrote:
Thanks, this is great news!
It was quite bad to hit disk limits and gluster to be oblivious to it and rebalancing is slow and taxing.
Any estimate in which version this feature will land?
If everything goes by plan, I think the 3.6 release should be built with
this patch.
- Varun Shastry
Lucian
--
Sent from the Delta quadrant using Borg technology!
Nux!
www.nux.ro
----- Original Message -----
From: "Varun Shastry" <vshastry@xxxxxxxxxx>
To: "Nux!" <nux@xxxxxxxxx>, gluster-devel@xxxxxxxxxxx
Sent: Friday, 27 June, 2014 12:41:46 PM
Subject: Re: Awareness of the disk space available on nodes
On Friday 27 June 2014 04:02 PM, Nux! wrote:
Hi,
In 3.3/3.4 we had a problem with distributed volumes having problems when a
brick/node would fill up and GlusterFS just failing to write down any more
files on it, not trying to write them down on another node.
I was told this is not a bug and it happens because of the hashing used to
distribute the files (and no checks being done on disk space).
My question is, has any of the above changed in 3.5?
The patch addressing this is up for review at
http://review.gluster.org/#/c/8093/. I don't think any other patch (at
least partially) addressing this issue is merged in 3.5.
- Varun Shastry
Thanks
Lucian
--
Sent from the Delta quadrant using Borg technology!
Nux!
www.nux.ro
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
http://supercolony.gluster.org/mailman/listinfo/gluster-devel
_______________________________________________
Gluster-devel mailing list
Gluster-devel@xxxxxxxxxxx
http://supercolony.gluster.org/mailman/listinfo/gluster-devel
