Proxmox noob here - I've been asked to setup a 3 node PVE 8.1.10 cluster specifically to investigate options for monitoring with our NMS.
First install, I hit the (apparently) well known CEPH v18.2 (Reef) dashboard compatibility problem, and subsequently rebuilt the cluster and installed CEPH v17.2 (Quincy). I can now start the CEPH dashboard, but I'm unable to start the CEPH restful module (needed for the NMS), regardless of whether dashboard is enabled or not.
Tying to enable restful from the cli gives the following error -
Error ENOENT: module 'restful' reports that it cannot run on the active manager daemon: PyO3 modules may only be initialized once per interpreter process (pass --force to force enablement).
Running with the --force parameter kills the dashboard module and produces a HEALTH_WARN state in the Proxmox CEPH status, with the PyO3 error shown in the summary window.
I found a thread in the Proxmox forums that seemed to imply an update shortly after Christmas fixed the problem, but I've installed all available (non-subscription) updates and the problem persists.
Ideally, I'd like to have both CEPH dashboard and restful working, so I can compare results from both tools. But I MUST have restful available for the NMS integration, even if it means killing dashboard.
Does anyone know how I can achieve this please?
First install, I hit the (apparently) well known CEPH v18.2 (Reef) dashboard compatibility problem, and subsequently rebuilt the cluster and installed CEPH v17.2 (Quincy). I can now start the CEPH dashboard, but I'm unable to start the CEPH restful module (needed for the NMS), regardless of whether dashboard is enabled or not.
Tying to enable restful from the cli gives the following error -
Error ENOENT: module 'restful' reports that it cannot run on the active manager daemon: PyO3 modules may only be initialized once per interpreter process (pass --force to force enablement).
Running with the --force parameter kills the dashboard module and produces a HEALTH_WARN state in the Proxmox CEPH status, with the PyO3 error shown in the summary window.
I found a thread in the Proxmox forums that seemed to imply an update shortly after Christmas fixed the problem, but I've installed all available (non-subscription) updates and the problem persists.
Ideally, I'd like to have both CEPH dashboard and restful working, so I can compare results from both tools. But I MUST have restful available for the NMS integration, even if it means killing dashboard.
Does anyone know how I can achieve this please?