Proxmox Datacenter Manager - First Alpha Release

Would be nice, to set confiured hosts offline in the administration., so there are no timeouts while maintenance
 
hi i updated to 1.9 today and noticed that my pdm can not shutdown via pdm gui or via the host node gui with the shutdown button. it can only shutdown via shell with the shutdown command.
 
Been playing around with this for a couple of days. Very impressed so far, considering this is only an alpha release. I'm not sure whether this has been fully implemented yet, but I tried to create a new user in "pdm" realm to avoid using root user. Creating the user works great, however I cannot seem to assign permissions to this user, so even though I can log on, I am getting all sorts of errors as the user is not authorised to do anything.
 
Guys, I have 3 clusters and I would like to know if with PDM I can create a user who can only see 3 VMs from different clusters? If not, are there any plans for this?
 
Is full PVE API support planned, so it can truly be used as a replacement for clustering (without HA of course)? My main use case would be to have one central server to make API calls to for all of my nodes, but we'd need full PVE API support. Creating VMs, statistics, editing hardware / throttling, etc.
 
Hello to the community and to the PROXMOX team,

I would like to understand how the values under Remotes => Allocation => CPU Cores assigned and Memory assigned come about?

I believe that the first area is the CPUs or memories that all VMs have together and the second area is all the physical CPUs or memories that all nodes have together, but what is the third area, which is in brackets, for a value?
 
Bu66as said:
Hello to the community and to the PROXMOX team,

I would like to understand how the values under Remotes => Allocation => CPU Cores assigned and Memory assigned come about?

I believe that the first area is the CPUs or memories that all VMs have together and the second area is all the physical CPUs or memories that all nodes have together, but what is the third area, which is in brackets, for a value?
Click to expand...
Hi, I would belive that this in the brackets is all, what you have configured. 37 running / 8 physical (75 total configured).
So my example had 37 cores running. But I try and test here and there with VMs and Software. So I have in total 75 cores assign to VMs, bt only 37 running at the moment.
 
I have discovered a bug. If I use my PDM Session from yesterday and want to sign in again, with TOTP too, then I will be redirectet to the login screen. I have to reload the entire site with new cookie.
 
husskii said:
hi i updated to 1.9 today and noticed that my pdm can not shutdown via pdm gui or via the host node gui with the shutdown button. it can only shutdown via shell with the shutdown command.
Click to expand...
Hi regarding this comment i made a couple days ago, it seems to of worked properly after installing qemu-agent on pdm.
 
Using CEPH I got migration to work, I have it running as a VM on PVE, and tested completely removing the VM and reinstalling, it kept saying the cert already existed and would not complete the install, until I removed the cert from the server...just a head up. Is there a way to have CPU temps on there???
 
  • Like
Reactions: husskii
Can I suggest being able to control the shutdown, reboot etc nodes via Datacenter Manager. We can do this with vms and lxc but I noticed its not possible with the actual node.
Also would be good to have more stats and monitoring, such as cpu temp, network usage etc..

Thanks again for the great work
 
  • Like
Reactions: wavejumper
Is it only me, but after I upgrade from 0.1.8 to 0.1.9 (both PDM and PDM Client) I get following API error on the "Notes" and "Administration" (incl. "Shell") sections:

Code: 
Load failed
api error (status = 400 Bad Request): error trying to connect: No such file or directory (os error 2)

Even worse, once I log out, I cannot log into PDM anymore as I receive this HTTP error:

Code: 
HTTP status 400

Only chance is to restore a previous backup of the PDM (0.1.8) VM and continue using 0.1.8 version. Error is in my environment (PDM is a dedicated VM on my PVE 8.3.2 Test Server) always reproducible.

EDIT:
Upgrade from 0.1.8 to 0.1.10 worked now flawlessly. Or just reboot as mentioned here. FYI
 

Attachments

  • pdm_-_Proxmox_Datacenter_Manager.png
    pdm_-_Proxmox_Datacenter_Manager.png
    27.9 KB · Views: 12
  • pdm1_-_Proxmox_Datacenter_Manager.png
    pdm1_-_Proxmox_Datacenter_Manager.png
    37.1 KB · Views: 12
Last edited:
Initially congratulations, already as Alpha it seems to be an excellent product.
I'm trying 0.1.9 and I noticed a strange behavior, if one of the remote servers becomes unreachable the interface becomes very slow and at syslog level it reports the impossibility of reaching the node xxx
Since PDM queries the individual hosts if a host becomes unreachable due to errors or routing problems it is possible that it becomes a frequent problem.
If instead the nodes/clusters were to register on PDM the server could "limit" itself to notifying the failure to receive data from the node XXX without continuing to try to connect and freeing up resources for the management of the other servers.
However, the centralized management of multiple hosts/clusters is certainly a clear step forward to multi-datacenter management
 
CoolWolf said:
Is it only me, but after I upgrade from 0.1.8 to 0.1.9 (both PDM and PDM Client) I get following API error on the "Notes" and "Administration" (incl. "Shell") sections:

Code: 
Load failed
api error (status = 400 Bad Request): error trying to connect: No such file or directory (os error 2)

Even worse, once I log out, I cannot log into PDM anymore as I receive this HTTP error:

Code: 
HTTP status 400

Only chance is to restore a previous backup of the PDM (0.1.8) VM and continue using 0.1.8 version. Error is in my environment (PDM is a dedicated VM on my PVE 8.3.2 Test Server) always reproducible.
Click to expand...
had the same issue after the update, just send the "reboot" from PVE and wait for it to come back up, should work after it comes back up.

Did not investigate the issue since reboot fixed it, but I assume some service failed to start back up properly after the update.
 
  • Like
Reactions: CoolWolf
husskii said:
Can I suggest being able to control the shutdown, reboot etc nodes via Datacenter Manager. We can do this with vms and lxc but I noticed its not possible with the actual node.
Also would be good to have more stats and monitoring, such as cpu temp, network usage etc..

Thanks again for the great work
Click to expand...
I think Wake on LAN support would also be great to add to PDM. So that we can power on a node that is currently off, like can be done in PVE clusters.

Thank you.
 
One node in a cluster of two nodes. One node shows proper statistic but the other does not for any VM's. Clicking on the node itself gives this error message:
api error (status = 400 Bad Request): api error (status = 595 <unknown status code>):
See attached screendump.2024-12-29_21-40.png
 
You must log in or register to reply here.
Top Bottom