Description of problem: Deployed OSP17 with STF. Deploy Yaml files included Collectd ExtraConfig configuration for virt plugin which suppose to provide instances names when showing them in promethues and grafana. collectd::plugin::virt::hostname_format: metadata uuid hostname collectd::plugin::virt::plugin_instance_format: metadata but name portion of created instances in OSP17 not included. For example: OSP16.2 instances show: plugin_instance="workload_instance_0" OSP17 instances show: plugin_instance="base"
Please provide more info. What is in the template and what is in the config? Are you saying this is a regression in puppet-collectd (i.e. the config is different)?
please use quotes around the parameters. collectd::plugin::virt::hostname_format: "metadata uuid hostname"
changing component from puppet-collectd to collectd, there was no change in puppet-collectd in the virt plugin for the last 2 years.
(In reply to Matthias Runge from comment #2) > Please provide more info. What is in the template and what is in the config? > Are you saying this is a regression in puppet-collectd (i.e. the config is > different)? So I am using identical templates for osp16.2 and osp17 (except clouds names) Both templates have the same lines: collectd::plugin::virt::hostname_format: metadata uuid hostname collectd::plugin::virt::plugin_instance_format: metadata plugin_instance_format: metadata will enable representation of created instances by their actual name in Prometheus e.g. you will see in prometheus VMs name like: plugin_instance="leonid" In OSP17 this doesn't work and I see generic plugin_instance="base" for all instances created in OSP17.
I was thinking this might be related to https://bugzilla.redhat.com/show_bug.cgi?id=1954103 but that covered the PluginInstance configuration, not the Hostname configuration. Please provide the packaged versions in 16.2 vs 17.0 for the puppet-collectd and collectd versions to determine if there is a version regression / non-bump. Will require additional investigation likely so we'll need to get this assigned to a developer to review code unless QE will perform the code walk and investigation.
(In reply to Leif Madsen from comment #8) > I was thinking this might be related to > https://bugzilla.redhat.com/show_bug.cgi?id=1954103 but that covered the > PluginInstance configuration, not the Hostname configuration. > > Please provide the packaged versions in 16.2 vs 17.0 for the puppet-collectd > and collectd versions to determine if there is a version regression / > non-bump. > > Will require additional investigation likely so we'll need to get this > assigned to a developer to review code unless QE will perform the code walk > and investigation. OSP 16.2: puppet-collectd-12.0.1-2.20220104100218.4686e16.el8ost.noarch collectd-5.11.0-10.el8ost.1.x86_64 OSP17: puppet-collectd-13.0.1-0.20220511160845.ad138a7.el9ost.noarch collectd-5.12.0-7.el9ost.x86_64
Tried to reproduce here, I see in the collectd log: libvirt: Domain Config error : metadata not found: Requested metadata element is not present Leonid, can you confirm?
Workaround? Set HostnameMetadataNS "http://openstack.org/xmlns/libvirt/nova/1.1" in the collectd virt plugin config. Unfortunately, that setting is not exposed in puppet-collectd
https://github.com/collectd/collectd/pull/4011
Fixed plugin_instance="workload_instance_1", service="default-osp17-coll-meter"
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (Release of components for Red Hat OpenStack Platform 17.0 (Wallaby)), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHEA-2022:6543