exporter queries, prometheus config file and dashboards for container deployments #188
Conversation
pmereddy1
changed the title
There was a problem hiding this comment.
One thing of note that Pramodh mentioned to me is that he had to manually modify these dashboards so that they could be imported via the Grafana API. We've had huge issues trying to make dashboards consistently compatible with the API import before and Grafana itself does not provide a means to export dashboards that are compatible with their import API. As of version 5.x of Grafana, we've been using automated provisioning and that is the method suggested from Grafana themselves for automated deployment.
If we can make sure that the method that these dashboards are installed in the containers uses the provisioning system, that will go a long way to ensuring the dashboards can be installed in a consistent manner for the foreseeable future.
@pmereddy1 For now, can you just upload the versions of the dashboard that are straight exports so we can try and see if we can get the provisioning system working with however this is going to be packaged? Please keep your modified ones around in case they are needed, tho.
…ds with new names, remove invalid panel links
… deployments (#188) (#190) * exporter queries, prometheus config file and dashboards * same dashboards. But exported from Grafana GUI * fix file permissions * add links and fix units on network panel * rename the 4 golden signals dashboard and add disk info to the saturation panel * change metric names from ccp_container to ccp_nodemx, update dashboards with new names, remove invalid panel links * rename queries_containers.yml to queries_nodemx.yml * fix the description of process count metric * exported dashboards to use externally Co-authored-by: Pramodh Mereddy <[email protected]>
* exporter queries, prometheus config file and dashboards for container deployments (#188) * exporter queries, prometheus config file and dashboards * same dashboards. But exported from Grafana GUI * fix file permissions * add links and fix units on network panel * rename the 4 golden signals dashboard and add disk info to the saturation panel * change metric names from ccp_container to ccp_nodemx, update dashboards with new names, remove invalid panel links * rename queries_containers.yml to queries_nodemx.yml * fix the description of process count metric * exported dashboards to use externally * Container related updates to Prometheus Configs (#191) * Update Prometheus Job Name and Source Label The 'Crunchy Collect' container has been moved from the Crunchy Containers project to the Postgres Operator project and renamed to 'Crunchy Postgres Exporter'. This updates the Prometheus config file for container deployments to match the new container name. * Update queries_nodemx.yml File Comment Updates the comment beginning and ending comments from referring to 'pod_metrics.yml' to the current name, 'queries_nodemx.yml'. Co-authored-by: Pramodh Mereddy <[email protected]> Co-authored-by: tjmoore4 <[email protected]>
Description
NOTE:
Fixes (issue) #
Depends on (issue) #
Type of change
Please check all options that are relevant
How Has This Been Tested?
Tested Configuration:
Tested with playbook(s):
Checklist: