Hello,
There are times when we need to address issues related to noisy guests by storage. For instance, if we encounter slow storage and need to identify which VM or LXC is utilizing the majority of the resources.
I have created some per-cluster/node dashboards in Grafana utilizing the default Proxmox VE metrics exporter for monitoring guests' CPU and memory consumption. However, when it comes to storage, I am struggling to find an efficient way to display only specific machines (represented by disks on specific storage).
Theoretically, I could achieve this by creating a variable with a list of disks via API requests using the marcusollson-json-datasource. However, I think that this approach may result in a high volume of API requests being executed each time the dashboard is opened or the filter is changed.
Maybe somebody already did something like this already?
Thanks.
There are times when we need to address issues related to noisy guests by storage. For instance, if we encounter slow storage and need to identify which VM or LXC is utilizing the majority of the resources.
I have created some per-cluster/node dashboards in Grafana utilizing the default Proxmox VE metrics exporter for monitoring guests' CPU and memory consumption. However, when it comes to storage, I am struggling to find an efficient way to display only specific machines (represented by disks on specific storage).
Theoretically, I could achieve this by creating a variable with a list of disks via API requests using the marcusollson-json-datasource. However, I think that this approach may result in a high volume of API requests being executed each time the dashboard is opened or the filter is changed.
Maybe somebody already did something like this already?
Thanks.