QuestionQ40

VerifiedAnalyze and Optimize Storage Performance

An administrator must ensure that an RF3 cluster can withstand the failure of two entire racks without adversely affecting performance. The current cluster configuration is as follows:

  • 30 All Flash Nodes, distributed as 10 nodes per rack across three 42U racks
  • Each node is configured with 20 TB of usable all-flash storage (cluster total: 600 TB usable)
  • Current cluster storage utilization is 500 TB
  • Storage containers have Erasure Coding enabled

Which configuration changes should be made to ensure the cluster meets these requirements?

Explanation

A 60-node cluster distributed evenly across six racks places 10 nodes in each rack. If two racks fail, 40 nodes remain, which exceeds the original 30-node operating footprint and therefore preserves node-based performance capacity. The surviving nodes also provide 800 TB of stated usable storage capacity, enough for the current 500 TB utilization. Six racks provide the required rack-level distribution for tolerating two complete rack failures.

Community Discussion

No comments yet. Be the first to start the discussion!