I'm having a weird issue where any connection to the proxmox nodes is disconnected every couple of minutes.
If i for example, select a node, open up the console via the web gui, after a couple of minutes, it will disconnect and I will have to close the pop up window and reopen it.
This also happens with SSH sessions which is making it difficult to transfer large files, as well as the console view for virtual machines. I have to constantly close and reopen them.
Here's the rub, if i ping the hosts, pings never drop or go outside of an acceptable value, they're very stable. If I RDP to one of the virtual machines running on the node, that connection never drops. IE so if i RDP to a machine, and on the same machine, open the console view, the console view will freeze and need reopened, but via rdp i never notice anything happen. This RDP happens through the same physical interface that the proxmox management is through, though it is a different vlan.
The general setup of the cluster is as follows:
3 vm only nodes, 3 ceph nodes that can run vm's, but are basically reserved for ceph
1 private network for ceph that just the ceph nodes are on (10gb)
1 public network for ceph that all 6 nodes are on (10gb)
1 cluster network that all 6 nodes are on for just cluster comms (1gb)
1 public network for management, this interface also has a vlan on a bridge for virtual machines to be attached to (1gb)
If i for example, select a node, open up the console via the web gui, after a couple of minutes, it will disconnect and I will have to close the pop up window and reopen it.
This also happens with SSH sessions which is making it difficult to transfer large files, as well as the console view for virtual machines. I have to constantly close and reopen them.
Here's the rub, if i ping the hosts, pings never drop or go outside of an acceptable value, they're very stable. If I RDP to one of the virtual machines running on the node, that connection never drops. IE so if i RDP to a machine, and on the same machine, open the console view, the console view will freeze and need reopened, but via rdp i never notice anything happen. This RDP happens through the same physical interface that the proxmox management is through, though it is a different vlan.
The general setup of the cluster is as follows:
3 vm only nodes, 3 ceph nodes that can run vm's, but are basically reserved for ceph
1 private network for ceph that just the ceph nodes are on (10gb)
1 public network for ceph that all 6 nodes are on (10gb)
1 cluster network that all 6 nodes are on for just cluster comms (1gb)
1 public network for management, this interface also has a vlan on a bridge for virtual machines to be attached to (1gb)