Replying to my own email here, but the problem is that the Ansible playbooks are removing metricsPublicURL and loggingPublicURL from master-config.yaml on the masters. This is happening even though I do *not* comment out the openshift_hosted_logging_master_public_url and openshift_hosted_metrics_master_public_url options in the Ansible hosts file.
Could this be a bug? On Wed, Jan 25, 2017 at 10:13 AM, Josh Baird <joshba...@gmail.com> wrote: > Hi all, > > Facing an odd situation here and was hoping for some feedback. > > I'm trying to stand up two OCP 3.3 HA environments (3/3/4) with logging > and metrics. I'm installing metrics from the Ansible playbooks like so: > > # metrics > openshift_hosted_metrics_deploy=true > openshift_hosted_metrics_storage_kind=nfs > openshift_hosted_metrics_storage_access_modes=['ReadWriteOnce'] > openshift_hosted_metrics_storage_host=fc-cifs03.corp.follett.com > openshift_hosted_metrics_storage_nfs_directory=/OCPQA_INFRA01 > openshift_hosted_metrics_storage_volume_name=metrics > openshift_hosted_metrics_storage_volume_size=75Gi > openshift_hosted_metrics_public_url=https://hawkular- > metrics.qa.ocp.domain.com/hawkular/metrics > > This results in a successful metrics installation. Metrics are displayed > for each pod on the overview page, etc. > > Next, I want to go back and add logging, so I comment out the metrics > stuff from the 'hosts' file, and add the logging config: > > # logging > openshift_hosted_logging_deploy=true > openshift_hosted_logging_storage_kind=nfs > openshift_hosted_logging_storage_access_modes=['ReadWriteOnce'] > openshift_hosted_logging_storage_host=<nfsserver> > openshift_hosted_logging_storage_nfs_directory=<nfsexport> > openshift_hosted_logging_storage_volume_name=logging > openshift_hosted_logging_storage_volume_size=150Gi > > After a successful playbook run, the logging pods come up fine (not sure > if they are totally operational), but something is causing my metrics data > to vanish. Metrics stats are no longer visibly on the pod overview page or > in the pod details page. I don't see any errors on any of the logging > and/or metrics pods. I have been able to consistently reproduce this in > multiple environments. > > Any ideas on how to troubleshoot this? > > Thanks. > >
_______________________________________________ users mailing list users@lists.openshift.redhat.com http://lists.openshift.redhat.com/openshiftmm/listinfo/users