System Dashboard (Telegraf + Netdata)

System Dashboard: Telegraf + Netdata

Based on https://grafana.com/grafana/dashboards/928, with tweaks.

The first page is a condensed system summary, including:

• CPU usage by core in single graph
• Memory usage (using the Committed_AS and Available values, which are more useful for overviews)
• Disk I/O for all disks in a single graph
• Network usage
• Linux Pressure Stall Information (PSI), collected by Netdata

Tweaks to the original dashboard:

• Display gaps for missing data in graphs: https://github.com/influxdata/influxdb/issues/7185#issuecomment-349024751
• Use $__interval instead of the deprecated $interval
• Avoid NON_NEGATIVE_DERIVATIVE() where not needed
• Singlestat panels that show guages are migrated to Guage panels

The influxdb datasource variables $telegraf and $netdata are hidden by default. To change them after importing the dashboard, open the dashboard settings, go to the variables section, click on the variable, change the “hide” selection box to the empty value, and click update. When you return to the dashboard, the variable will be modifiable at the top of the dashboard next to the other variables.