There is still ongoing discussion, as the patches are in an early stage so which storage types are going to be supported cannot be answered just yet, see e.g. https://lists.proxmox.com/pipermail/pve-devel/2024-January/061519.html
Hi,
the storage where VM images are on doesn't matter. The storage for the fleecing images needs to be capable of thin volumes or sparse files and support discard. LVM-thin does both and is very likely going to be supported. Storages without those features will require much more space usage and we're still discussion whether we should simply prohibit those or only recommend/warn against (some users could might still want to use such a storage if the space usage is not an issue for them).
the storage where VM images are on doesn't matter. The storage for the fleecing images needs to be capable of thin volumes or sparse files and support discard. LVM-thin does both and is very likely going to be supported. Storages without those features will require much more space usage and we're still discussion whether we should simply prohibit those or only recommend/warn against (some users could might still want to use such a storage if the space usage is not an issue for them).
HI,
i am just testing fleecing, and may i offer an opinion, i think it should be possible to use it on non thin-provisioned disks (for my taste a warning is fine)
Cheers
Soeren
i am just testing fleecing, and may i offer an opinion, i think it should be possible to use it on non thin-provisioned disks (for my taste a warning is fine)
Cheers
Soeren
We are using ceph and NFS. And the servers are fully loaded with disks already, therefore no space for a new local disk to use as fleece disk.
But we have tons of space on the ceph and nfs. THe fleece disk on ceph works just fine, even though we have VMs with disks that are several TB
But we have tons of space on the ceph and nfs. THe fleece disk on ceph works just fine, even though we have VMs with disks that are several TB
I'm also currently testing it with with ceph as fleecing storage, with the vm on the same rbd storage too. (full nvme cluster)
(backup server is long distance with hdd)
So far, no problem, no crash. (So, I think It'll be able to finally migrate to PBS )
(backup server is long distance with hdd)
So far, no problem, no crash. (So, I think It'll be able to finally migrate to PBS )
Hi @dietmar.
We expect our customers to use the same Blockbridge storage for fleecing as the one that backs their virtual disks.
The protection they are looking for is from a slow PBS and/or network hiccups between PVE and PBS. The storage is fast and can handle all of the flows.
Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
Hi,
you can already configure any storage (which supports allocating images) as a fleecing storage. There is no restriction in the code regarding that.
Last edited: