Prometheus Targets configured via CMR are not named usefully

Bug #1943272 reported by Drew Freiberger
8
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Prometheus2 charm
New
Wishlist
Unassigned

Bug Description

When deploying prometheus2 charm into a separate model from the telegraf charm, for instance, the name of the exporter endpoint type is "remote-$UUID" instead of "telegraf". This makes investigating prometheus target endpoints tedious as you need to know the port numbers for each exporter (telegraf, p-ceph-exporter, p-openstack-exporter, p-libvirt-exporter, etc).

It would be very useful to get the proper remote application name through the CMR relation to exporters.

To reproduce, create two models, lma and workload, offer the

juju add-model workload
juju deploy cs:ubuntu
juju deploy cs:telegraf
juju offer telegraf:prometheus-client

juju add-model lma
juju deploy cs:prometheus2
juju add-relation prometheus2:target admin/workload.telegraf

juju status prometheus2 # to get the service IP and port
visit the prometheus2 endpoint and click on Status->Targets and observe that the telegraf client endpoints have a header of "remote-$some_uuid (1/1 up)". It would be helpful if this showed something like "workload.telegraf" or "$workload_model_uuid.telegraf".

Changed in charm-prometheus2:
importance: Undecided → Wishlist
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.