<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"><!-- P {margin-top:0;margin-bottom:0;} --></style>
</head>
<body dir="ltr">
<div id="divtagdefaultwrapper" style="font-size:12pt;color:#000000;background-color:#FFFFFF;font-family:'Lucida Console', Monaco, monospace;">
<p></p>
<div>Hello,</div>
<div><br>
</div>
<div>We use a Ceph cluster for Nova (Glance and Cinder as well) and over time,</div>
<div>more <span style="font-size: 12pt;">and more data is stored there. We can't keep the cluster so big because of</span><span style="font-size: 12pt;"> </span></div>
<div>Ceph's limitations. Sooner or later it needs to be closed for adding new </div>
<div>instances, images and volumes. Not to mention it's a big failure domain.</div>
<div><br>
</div>
<div>How do you handle this issue?</div>
<div>What is your strategy to divide Ceph clusters between compute nodes?</div>
<div>How do you solve VM snapshot placement and migration issues then</div>
<div>(snapshots will be left on older Ceph)?</div>
<div><br>
</div>
<div>We've been thinking about features like: dynamic Ceph configuration</div>
<div>(not static like in nova.conf) in Nova, pinning instances to a Ceph cluster etc.</div>
<div>What do you think about that?</div>
<div><br>
</div>
<br>
<p></p>
</div>
</body>
</html>