> In hyper-converged deployments, the node performing the backup is sourcing > ((nodes-1)/(nodes))*bytes) of backup data (i.e., ingress traffic) and then > sending 1*bytes to PBS (i.e., egress traffic). If PBS were to pull the data > from the nodes directly, the maximum load on any one host would be > (1/nodes)*bytes of egress traffic only... that's a considerable improvement!
I guess it would be possible to write a tool like proxmox-backup-client that pull ceph backups directly from PBS. Or extend the backup protokoll allowing direct storage access. But this is a considerable amount of development, and needs much more configuration/setup than the current approach. But patches are always welcome... Also, it is not clear to me how we can implement a "backup provider API" if we add such optimizations? And yes, network traffic would be reduced. But IMHO it is easier to add a dedicated network card for the backup server (if the network is the limiting factor). With this setup, the maximum load on the ceph network is (1/nodes)*bytess of egress traffic only. The backup traffic is on the dedicated backup net. _______________________________________________ pve-devel mailing list pve-devel@lists.proxmox.com https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
