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
Blockscope.net would like to provide in this issue, our vision from our experience implementing and using grafana dashboards, in case it is useful for creating next versions of the dashboard.
1.- In our experience hardware and system usage metrics are nice to have, but usually node operators rely on their own "standard" grafana dashboards. These dashboards are already configured by node operators with custom alerts and metrics. So hardware metrics are good to have but not mandatory IMO. Below the ones we use for our servers.
2.- Thus we think most important metrics are the ones related to consensus, blocks, transactions, rounds, epochs, and other interesting Supra network related metrics. Creating widgets with this information could expose sincronization, connectivity, network congestion, etc. issues. This would allow also to implement specific alerts for Supra network on the Grafana dashboard itself. Below a dashboard we've built for Sui network that exposes interesting network metrics.
3.- Probably would be good to publish / OS the grafana dashboard code for node operators that already have their own prometheus/grafana stack in place (most of node operators do). So N.O. can tailor and modify the dashboard to their needs besides having an official one. Previous Sui dashboard below.
4.- Another idea, would be to implement a network generic metrics dashboard in https://monitoring.services.supra.com and accessible for NO to check issues that may arise, or to correlate observations looking to other nodes charts (such as DC or geolocation related issues). Below an example of this kind of dashboard.
I've seen that these consensus/network metrics are already in the todo list as "Application level" metrics (https://github.com/orgs/Entropy-Foundation/projects/13/views/1), so waiting for those to be implemented. We'll be willing to help build the next dashboard version to include these metrics, and also explore the possibility to build a monitoring CLI tool around these metrics as we've done in other networks.
Hope this information can help you on evolving this dashboard, and feel free to discuss anything related to it and our vision.
The text was updated successfully, but these errors were encountered:
Blockscope.net would like to provide in this issue, our vision from our experience implementing and using grafana dashboards, in case it is useful for creating next versions of the dashboard.
https://grafana.com/grafana/dashboards/15172-node-exporter-for-prometheus-dashboard-based-on-11074/
https://grafana.com/grafana/dashboards/1860-node-exporter-full/
https://gitlab.com/blockscope-net/sui-tools/-/blob/main/README.md?ref_type=heads
I've seen that these consensus/network metrics are already in the todo list as "Application level" metrics (https://github.com/orgs/Entropy-Foundation/projects/13/views/1), so waiting for those to be implemented. We'll be willing to help build the next dashboard version to include these metrics, and also explore the possibility to build a monitoring CLI tool around these metrics as we've done in other networks.
Hope this information can help you on evolving this dashboard, and feel free to discuss anything related to it and our vision.
The text was updated successfully, but these errors were encountered: