Red Hat Training

A Red Hat training course is available for Red Hat OpenStack Platform

第2章 Configuration recommendations for the Telemetry service

Because the Red Hat OpenStack Platform (RHOSP) Telemetry service is CPU-intensive, telemetry is not enabled by default in RHOSP 13. However, by following these deployment recommendations, you can avoid performance degradation if you enable telemetry.

These procedures—​one for a small, test overcloud and one for a large, production overcloud—​contain recommendations that maximize Telemetry service performance.

2.1. Configuring the Telemetry service on a small, test overcloud

When you enable the Red Hat OpenStack Platform (RHOSP) Telemetry service on small, test overclouds, you can improve its performance by using a file back end.

Prerequisites

  • The overcloud deployment on which you are configuring the Telemetry service is not a production system.
  • The overcloud is a small deployment that supports fewer than 100 instances, with a maximum of 12 physical cores on each Controller node, or 24 cores with hyperthreading enabled.
  • The overcloud deployment has high availability disabled.

Procedure

  1. Add the following to parameter_defaults in your /usr/share/openstack-tripleo-heat-templates/environments/enable-legacy-telemetry.yaml environment file: 

    parameter_defaults:
  GnocchiBackend: file

  2. Add the enable-legacy-telemetry.yaml file to your openstack overcloud deploy command: 

    openstack overcloud deploy \
-e /home/stack/environment.yaml \
-e /usr/share/openstack-tripleo-heat-templates/environments/enable-legacy-telemetry.yaml \
[...]

Additional resources