QuestionQ44

VerifiedIntroduction to Performance in Clustered Data ONTAP

You have an ONTAP two-node cluster. A read/write test workload against an SMB share is slower than expected. You run the same test again, this time monitoring storage by using a QoS command.

Question Image

Based on the exhibit, what is the problem?

Explanation

The high latency is concentrated in NVRAM transfer, not in the disk, client network, cluster, data-processing, or QoS components. In an ONTAP HA pair, a write is logged to NVRAM on both nodes before acknowledgement, so elevated NVRAM-transfer latency indicates a problem on the HA interconnect used to mirror the NVRAM log.

Learn more

Community Discussion

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