Hello,
We’ve been running Grafana with proxy access to Zabbix data source for some while. The server has a 4 core CPU, 8GB memory and 120GB disk space with 1 Gbps network connection. We usually experiences delays from minutes to 10+ minutes while loading a dashboard.
What could be the bottleneck(s) that cause of the delay?
Would sizing up the hardware resolve the issue?
Would you please share your hardware configuration?
Thanks very much.
When you say “while loading a dashboard”, do you mean loading the panel data or the actual dashboard?
Grafana loads dashboards within 10-500ms so if you see longer times it is because you have template variables that are set to refresh on “dashboard load”, grafana will wait to initialise the dashboard panels until the template variable query completes.
Grafana requires very little hardware as the bottleneck is the time series database backend. Grafana can run very quickly on 250mb RAM (might need a little more if you have a lot of alert rules ) and a single core.
What needs beefy hardware is your TSDB (Zabbix)
Thanks very much for the clarification on the root cause of the delay and the reference size of Grafana.