Hi,
Not a problem but a need of advices about Dashboard to monitor Ceph.
We use Ceph, installed with Proxmox GUI.
Like the Proxmox/influxDB/Grafana stack, I'm looking to aggregate Ceph monitoring data in Prometheus (for what i've seen, one of the best options) and then display thoses data with Grafana.
But connecting Ceph with Prometheus can be done 2 ways (at least) :
--> Activate Prometheus module in Mgr
- Simple to activate/configure
- Uses ressources from Mgr process
--> Configure Ceph_Exporter
- Not as simple
- Independant of Mgr from ressources perspectives
So my questions :
- Does Prometheus module consume lots of ressources from Mgr or is it a problem for large scale Ceph clusters ?
- What is a large scale cluster in this situation ? (Some talk about Thousands of OSD, we have 3 nodes, ~50 OSDs
- Does Ceph_Exporter as more advantages that I doesn't seen
Feel free to correct me if I'm wrong on some points
Not a problem but a need of advices about Dashboard to monitor Ceph.
We use Ceph, installed with Proxmox GUI.
Like the Proxmox/influxDB/Grafana stack, I'm looking to aggregate Ceph monitoring data in Prometheus (for what i've seen, one of the best options) and then display thoses data with Grafana.
But connecting Ceph with Prometheus can be done 2 ways (at least) :
--> Activate Prometheus module in Mgr
- Simple to activate/configure
- Uses ressources from Mgr process
--> Configure Ceph_Exporter
- Not as simple
- Independant of Mgr from ressources perspectives
So my questions :
- Does Prometheus module consume lots of ressources from Mgr or is it a problem for large scale Ceph clusters ?
- What is a large scale cluster in this situation ? (Some talk about Thousands of OSD, we have 3 nodes, ~50 OSDs
- Does Ceph_Exporter as more advantages that I doesn't seen
Feel free to correct me if I'm wrong on some points