i've created a bp to discuss whether moving the alarming into pipeline is feasible and can cover all the use cases for alarm. if we can find a solution that is a bit leaner than what we have and still provide same functionality coverage i don't see why we try it. it very well may be that what we have is the best solution. https://blueprints.launchpad.net/ceilometer/+spec/alarm-pipelines cheers, gordon chung openstack, ibm software standards -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20140311/9e81c851/attachment.html>