Vaughan; An absolute minimal Ceph cluster really needs to be 3 servers, and at that usable space should be 1/3 of raw space (see the archives of this mailing list for many discussions of why size=2 is bad). While it is possible to run other tasks on Ceph servers, memory utilization of Ceph processes can be quite large, so it's often discouraged, especially on memory constrained servers. Would it be feasible to acquire a system with sufficient RAM to run both VMs? I believe RBD can be cached, but I can't speak to how it's configured, or how well it works. I believe you would want a really fast drive (SSD) to store the cache on. Depending on your performance and storage volume needs, you might be able to get away with building a micro-cluster, based on ARM CPUs. Thank you, Dominic L. Hilsbos, MBA Director - Information Technology Perform Air International Inc. DHilsbos@PerformAir.com www.PerformAir.com -----Original Message----- From: Vaughan Beckwith [mailto:Vaughan.Beckwith@bluesphere.co.za] Sent: Tuesday, November 17, 2020 3:54 PM To: ceph-users@ceph.io Subject: [ceph-users] Accessing Ceph Storage Data via Ceph Block Storage Hi All, I'm not sure if this is the correct place to ask this question, I have tried the channels, but have received very little help there. I am currently very new to Ceph and am investigating it as to a possible replacement for a legacy application which use to provide us with replication. At the moment my company has three servers, two primary servers running Ubuntu and a backup server also running Ubuntu, the two primary servers each host a virtual machine, and it is these virtual machines that the office workers use for shared folder access, email and as a domain server, the office workers are not aware of the underlying linux servers. In the past the legacy software would replicate the running VM files on both primary servers to the backup server. The replication is done at the underlying linux host level and not from within the guest VMs. I was hoping that I could get Ceph to do this as well. From what I have read and I speak under correction, the best Ceph client type for this would be the block access, whereby I would then mount the block and start up the VMs. As I would be running the VMs, as per normal routine, would Ceph then have to retrieve the large VM files from the storage nodes across the lan and bring the data back to the client to run in the VM. Is there an option to cache certain parts of the data on certain clients? Also none of the primary servers as they currently stand have the capacity to run both VMs together, so each primary has a dedicated VM which it runs, the backup server currently keeps replicated copies of both VM images from each primary, the replication is provided by the legacy application. I'm also wondering if I need to get a fourth server, so I have 2 clients and 2 storage nodes. Any suggestions or help would be greatly appreciated. Yours sincerely Vaughan Beckwith Bluesphere Technologies BSC I.T. (Honours) vaughan.beckwith@bluesphere.co.za<mailto:vaughan.beckwith@bluesphere.co.za> Telephone: 011 675 6354 Fax: (011) 675 6423 _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io