On Fri, Oct 16, 2015 at 1:31 PM, Kapetanakis Giannis <bilias@edu.physics.uoc.gr> wrote:
On 16/10/15 01:01, Nir Soffer wrote:
I thought something like this might work:
node[1]: FC - ISCSI <-> node[2]: ISCSI - FC
I dont follow - how do you want to share your fc storage over iscsi?
And what are these nodes? Storage nodes? Hypervisors?
Nir
The initial idea was to share FC over IP network. All nodes are ovirt nodes (hypervisors).
One node see the FC storage share as a block device. Take this block device and share it over iscsi to node[2]. The node[2] will see the block device and create a new FC target to use it itself...
I know, its science fiction...
I still don't understand the problem you are trying to solve. Can you explain the network topology? - Do you have FC storage server, switch? - How many nodes to do you have with FC HBA? Do you want to add nodes without FC HBA, and you want to consume the FC storage? Or you want to add nodes with FC HBA, but you don't have an FC switch? Nir
After reading a bit on the subject another solution would be FCoE vn2vn but this probably requires hardware. Specific switches with DCB and FIP snooping which i don't have.
Any idea to extend FC to nodes without FC HBAs is welcome.
G