The cpu_util is a pain-point for us as well, we will unfortunately need to add that metric back to keep backward compatibility to our customers.

Best regards Tobias

On 9/9/19 11:37 AM, Blom, Merlin, NMU-OI wrote:

From Witek Bedyk witold.bedyk@suse.com on Re: [aodh] [heat] Stein: How to create alarms based on rate metrics like CPU utilization? Fr 16.08.2019 17:11 ' Hi all,

You can also collect `cpu.utilization_perc` metric with Monasca and trigger Heat auto-scaling as we demonstrated in the hands-on workshop at the last Summit in Denver.

Here the Heat template we've used [1]. You can find the workshop material here [2].

Cheers Witek

[1] https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_sjamgade_mon... [2] https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_sjamgade_mon... '

Cheers Merlin

-----Ursprüngliche Nachricht----- Von: Budai Laszlo laszlo.budai@gmail.com Gesendet: Freitag, 16. August 2019 18:10 An: OpenStack Discuss openstack-discuss@lists.openstack.org Betreff: [metrics] [telemetry] [stein] cpu_util

Hello all,

the release release announce of ceilometer rocky is deprecating the cpu_util and *.rate metrics "* cpu_util and *.rate meters are deprecated and will be removed in future release in favor of the Gnocchi rate calculation equivalent."

so we don't have them in Stein. Can you direct me to some document that describes how to achieve these with Gnocchi rate calculation?

Thank you, Laszlo

He guys, we are working to address this issue of Ceilometer. The first PR of a series that we are working on is the following: https://review.opendev.org/#/c/677031/. The idea is to leverage Ceilometer to enable operators/admins to create pollsters on the fly (without coding).

Reviews are welcome to help us get the first commit of the feature in. Also, feedback and suggestions for further improvements.

On Fri, Oct 18, 2019 at 12:23 AM Rico Lin rico.lin.guanyu@gmail.com wrote:

FYI, there's a new ML with similar topic: [1]

[1] http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010210.h...

On Mon, Sep 9, 2019 at 6:19 PM Tobias Urdin tobias.urdin@binero.se wrote:

...

The cpu_util is a pain-point for us as well, we will unfortunately need to add that metric back to keep backward compatibility to our customers.

Best regards Tobias

On 9/9/19 11:37 AM, Blom, Merlin, NMU-OI wrote:

...

From Witek Bedyk witold.bedyk@suse.com on Re: [aodh] [heat] Stein:

How to create alarms based on rate metrics like CPU utilization?

...

Fr 16.08.2019 17:11 ' Hi all,

You can also collect `cpu.utilization_perc` metric with Monasca and

trigger Heat auto-scaling as we demonstrated in the hands-on workshop at the last Summit in Denver.

...

Here the Heat template we've used [1]. You can find the workshop material here [2].

Cheers Witek

[1]

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_sjamgade_mon...

...

[2]

https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_sjamgade_mon...

...

'

Cheers Merlin

-----Ursprüngliche Nachricht----- Von: Budai Laszlo laszlo.budai@gmail.com Gesendet: Freitag, 16. August 2019 18:10 An: OpenStack Discuss openstack-discuss@lists.openstack.org Betreff: [metrics] [telemetry] [stein] cpu_util

Hello all,

the release release announce of ceilometer rocky is deprecating the

cpu_util and *.rate metrics

...

"* cpu_util and *.rate meters are deprecated and will be removed in future release in favor of the Gnocchi rate calculation equivalent."

so we don't have them in Stein. Can you direct me to some document that

describes how to achieve these with Gnocchi rate calculation?

...

Thank you, Laszlo

-- May The Force of OpenStack Be With You,

*Rico Lin*irc: ricolin