You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
State only the object ids as label to be scraped. I would welcome the addition of the clear-text values as labels in addition to the ids.
Problem or use case
After scraping a Syncthing server using Prometheus, I can display the metrics with tools like Grafana. Sadly, I can only display the unresolved ids of the folders / devices inside the UI, as no further information is available.
This leads to problems that alerts cannot contain easy-to-interepret descriptions - for admins to decide if they want to start investigation. In addition, admins cannot interpret e.g. if spikes in traffic are legitimate (as the device is only visible as ID) in Grafana, which makes anomaly-investigations harder.
Finally, with growing numbers of devices and folders connected to an instance (for me >40 folders / >8 devices) the metrics get more and more hard to read - pie charts containing statistics about specific folders in comparison to each other are just unreadable.
Here are some samples (blue = mask for privacy):
Alternatives or workarounds
Write an own exporter to scrape Syncthing-metrics and label them with the metadata available over the api.
Relabel all ids by hand in the Prometheus config?
The text was updated successfully, but these errors were encountered:
Feature description
Currently, metrics like:
State only the object ids as label to be scraped. I would welcome the addition of the clear-text values as labels in addition to the ids.
Problem or use case
After scraping a Syncthing server using Prometheus, I can display the metrics with tools like Grafana. Sadly, I can only display the unresolved ids of the folders / devices inside the UI, as no further information is available.
This leads to problems that alerts cannot contain easy-to-interepret descriptions - for admins to decide if they want to start investigation. In addition, admins cannot interpret e.g. if spikes in traffic are legitimate (as the device is only visible as ID) in Grafana, which makes anomaly-investigations harder.
Finally, with growing numbers of devices and folders connected to an instance (for me
>40
folders />8
devices) the metrics get more and more hard to read - pie charts containing statistics about specific folders in comparison to each other are just unreadable.Here are some samples (blue = mask for privacy):
Alternatives or workarounds
The text was updated successfully, but these errors were encountered: