Search results

Corosync should always run on its own NIC port. Check out the forum thread and benchmarks (incl. the 2018). https://forum.proxmox.com/threads/proxmox-ve-ceph-benchmark-2020-09-hyper-converged-with-nvme.76516/

Ceph 14.2.16 has been released, but the patch was not merged yet.

As said previously, the storage needs to be capable of doing snapshots. If not then you will need to use qcow2 image files on VM creation. https://pve.proxmox.com/pve-docs/chapter-pvesm.html Cifs/NFS use qcow2.

That won't work, since only a RBD/CephFS storage is configured. No ceph.conf is used then. But that doesn't say what the require-min-compat-client of the cluster is. We can only give support on our Ceph packages. A simple pveceph install will take care of it.

Did you create a VM/CT on it and start it? As the local-lvm is the default storage for a LVM based setup. I am not certain to why it doesn't get mapped. But I guess it may appear once you ran a VM/CT on it.

Please be aware that manual changes will get overwritten on package update. Please post the patch to pve-devel for further integration, if wished. https://pve.proxmox.com/wiki/Developer_Documentation

Ok 6.3-3. Any multipath/iscsi setups? And can you create a VM disk on that storage?

And what's the ceph osd get-require-min-compat-client? By default Debian ships with the Luminous client.

Depends on how big the version difference is. Nautilus can speak with Octopus (newer) and Luminous (older). And what permissions the user of the keyring has.

What ZFS version do the PVE and the Ubuntu server use? Can you please post a zfs get all <pool> and a zpool get all <pool>?

Wie gesagt, das macht alles der Installer für dich. Dort sind bereits zwei Storages eingerichtet, local und local-lvm. Wenn du jetzt einen Container oder eine VM erstellst, wählst du einfach local-lvm aus und dann wird dort das Disk Image für die VM erstellt.

What's the pveversion -v ?

That seems to be ok. The mapper might not have created the device node. The easiest to see if that persists, is to reboot the node.

To prevent locking or to make it happen? In either case it's ZFSs doing, not any case handling. But I can't eliminate the issue that in some constellations this might happen. The multipath output might show something like that and you blacklisted and removed multipath. But now, as we speak...

This can be done, and will limit the traffic to that VLAN. But the other network components need to know about the VLAN as well. And routing needs to be done between VLAN separated networks. But this sounds like the MAC address might be cached somewhere and traffic is not passed back to that MAC.

Weird. Can you post a get all from the pool and from the parent zfs dataset?

Proxmox VE is a full Linux distribution. Setting up a VPN is in most cases no issue.

From the top of my head. Try to move the CT to a different bridge and back. Re-check netmask, VLAN ID.

Well, the pve-zsync and the storage replication are a little different. You can specify on the disk config that it should be skipped by replication or not. Then please check that there is no send/receive running or stuck. Does a rename work? And are there any bookmarks?

Im Zweifel, einfach den Installer machen lassen. ;) Und anpassen lässt's sich in den meisten Fällen auch.