Hongbin & Lars, I spoke with Steve Hardy from Heat, who has been deep into improving resource groups. He indicated the problem Hongbin saw, where the resource group kills all Vms and then restarts them on a stack update to the count in a resource group is fixed in master and stable/juno.
See reviews: https://review.openstack.org/#/c/141820/ More important is this review https://review.openstack.org/#/c/131538/ Which implies a replace policy of auto is required for ports, or the vms could potentially be ripped out. A good thread about why that is needed: http://lists.openstack.org/pipermail/openstack-dev/2014-October/049376.html I think adding the replacement_policy:auto for ports should do the trick, but read through the thread and the last review to make your own conclusions. Thanks shardy for providing all the info in this email (I am just documenting) Regards -steve
__________________________________________________________________________ OpenStack Development Mailing List (not for usage questions) Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev