[StarCluster] StarCluster LoadBalancer

Adam adamnkraut at gmail.com
Wed Jan 23 10:44:58 EST 2013


Hi Sergio,

This is expected behavior for the loadbalancer. You don't have any queued
jobs. Your job is in running 'r' state. Also notice the load balancer
reports

>>> Cluster was modified less than 180 seconds ago
>>> Waiting for cluster to stabilize...

This means that it will not make any decision since the cluster is not
'stable' enough. You need to wait a bit longer. You can change the metrics
the loadbalancer uses from the command-line.

Hope this helps,
Adam

On Wed, Jan 23, 2013 at 8:39 AM, Sergio Mafra <sergiohmafra at gmail.com>wrote:

> Cluster was modified less than 180 seconds ago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/starcluster/attachments/20130123/5cf21202/attachment.htm


More information about the StarCluster mailing list