Hi, I would like to replace obsolete storage systems in production which are used for virtual machines under VMWare with KVM virtualization(proxmox) and SDS storage. I chose DRBD9 which seems more understandable for me than Ceph. The environment is around 150 VM and 40 TB storage.
1. I'm not sure about scalability(it goes from my lack of knowledges). Let's say I have 3 DRBD(3 replica) servers full of disks which serve for virtualization. Then, the capacity is over and I need to add more for next virtual machines. What should I add? Additional at least two servers? Or three with another quorum? What is the best approach(price/effectivity) in capacity sizing with regard of future painless capacity expanding. Yes, I could buy 24x10 TB SSD for each server and I would have a lot of capacity, but the price.. 2. If I would deploy 3 node drbd cluster (on of them quorum, no diskless mode) how many nodes could I loose with noninterupted storage services(it means available virtual machines). Two or only one? Thank you very much. Best regards, Vladislav
_______________________________________________ Star us on GITHUB: https://github.com/LINBIT drbd-user mailing list [email protected] https://lists.linbit.com/mailman/listinfo/drbd-user
