An administrator needs to make sure an RF2 cluster can survive a complete rack failure without negatively effecting workload performance. The current cluster configuration consists of the following:
✑ 30 nodes: distributed 10 nodes per rack across three 42U rack
✑ Each nodes is configured with 40TB usable storage all flash (Cluster Total 1.2 PB Usable)
✑ Current cluster utilization is 900TB storage
Which configuration changes should be made to make sure that the cluster meets the requirements?
To ensure the RF2 cluster can survive a complete rack failure without negatively affecting workload performance, the cluster must be able to maintain its storage needs and redundancy even if an entire rack is lost. The current cluster utilization is 900TB, and with each rack currently having 10 nodes, losing one rack would mean losing 400TB of storage. To maintain performance and avoid storage issues, the cluster requires additional nodes to compensate for this potential loss. Expanding the cluster by adding 9 nodes (3 additional nodes per rack) will increase the storage capacity adequately distributed across the racks and ensure redundancy and performance are maintained even in the event of a rack failure.
you need 50TB additional per rack (in order to get 450TB per rack), each node has 40TB so you need 2 more per rack, but 1 additional node (per rack) in order to get N+1.
What is the reason? 39 nodes?