Hi Guys,
Forgive me if this is a stupid question, but this is the first time I have touched LXC. I have used Proxmox for years but always used KVM, so LXC is a bit new to me.
My setup
- 3 Nodes
- All nodes use local storage
- All nodes are setup in a cluster
- All nodes are identical versions of PVE (Hardware varies a little)
I created a LXC container on Node01 with a 1TB vdisk using the raw (Seems raw is the only option?) and once I configured the container to my liking the raw vdisk reported ~16GB.
I proceeded to migrate the container to another node where I discovered the transfer was going to take ages (Its still running as a type this) which was unexpected.
As the raw vdisk is only ~16GB I expected the migration to only take a few minutes, however looking at the transfer log PVE is acting as its transferring a full 1TB vdisk.
Is this to be expected with the raw format?
Any options to get around this other than only growing the vdisk when you absolutely have to?
With KVM I always stuck with 2cow for its thin provisioning aspect and file management so I expected the transfer to be somewhat the same with LXC, guess I was mistaken?
Thanks!
Forgive me if this is a stupid question, but this is the first time I have touched LXC. I have used Proxmox for years but always used KVM, so LXC is a bit new to me.
My setup
- 3 Nodes
- All nodes use local storage
- All nodes are setup in a cluster
- All nodes are identical versions of PVE (Hardware varies a little)
I created a LXC container on Node01 with a 1TB vdisk using the raw (Seems raw is the only option?) and once I configured the container to my liking the raw vdisk reported ~16GB.
I proceeded to migrate the container to another node where I discovered the transfer was going to take ages (Its still running as a type this) which was unexpected.
As the raw vdisk is only ~16GB I expected the migration to only take a few minutes, however looking at the transfer log PVE is acting as its transferring a full 1TB vdisk.
Is this to be expected with the raw format?
Any options to get around this other than only growing the vdisk when you absolutely have to?
With KVM I always stuck with 2cow for its thin provisioning aspect and file management so I expected the transfer to be somewhat the same with LXC, guess I was mistaken?
Thanks!