<html><head></head><body><div style="font-family: Verdana;font-size: 12.0px;"><div>
<pre>Hi Nux,
ok it looks like this, but should gluster manage it by his self, therefore this situation should never happend?
thx
On 28.01.2014 09:09, Dragon wrote:
</pre>
<blockquote>Hi, i run a 3 brick glusterfs 3.4.0 distributed Cluster. The problem is that it run out of free space. While df -h on Client said 462G free i cant copy 30G File. Error is "no space left". I checked Option min-free-disk, but it is configured to 500MB. Than i checked the free inodes and on all 3 there are more than 11061089 inodes free. All 3 are online, on 1. are 10GB free, 2. 343MB and 3. 405GB free. While i copied i was connected via glusterfs mount to node1. I cant find the problem for this, need help thx. Sunny</blockquote>
Hi, I think it's trying to write that file on a brick that has little available space left, by what you wrote only one of the 3 bricks has enough space to accomodate your request. Lucian <span class="sig">-- Sent from the Delta quadrant using Borg technology! Nux!</span></div></div></body></html>