Open Stack

In Kilo, with the bug fix from 1326721, I can create an alarm like the
following with the "query" property. I use this alarm to monitor the health
of a server instance in the pool. In juno/stable, without the query
property, ceilometer alarms can only be created via heat with matching
metadata. This limits the metrics I can create alarms for, as not every
resource has metadata associated with it. Without this bug fix, I've been
unable to find a way to monitor the health of my server in Juno via heat.
I'd like thoughts on back porting 1326721 back to juno/stable, so this can
be work the same as it does in Kilo
https://review.openstack.org/#/c/146624/ .Thanks

gone_alarm:
    type: OS::Ceilometer::Alarm
    properties:
      description: Detect server being unresponsive
      repeat_actions: False
      meter_name: network.services.lb.member
      statistic: avg
      period: 600
      evaluation_periods: 1
      threshold: 1
      alarm_actions: [ {get_attr: [restart, AlarmUrl]} ]
      query:
      - field: resource_id
        op: eq
        value: {get_resource: member}
      comparison_operator: lt

  member:
    type: OS::Neutron::PoolMember
    properties:
      pool_id: {get_param: pool_id}
      address: {get_attr: [server_node, first_address]}
      protocol_port: { get_param: loadbalance_port }


Karolyn Chambers
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150121/ea250dce/attachment.html>