How to reduce the disk latency or data processing latency by volume move

Applies to

ONTAP 9

Description

This article describes how to migrate the busy volumes to less utilized nodes/aggregates to resolve the high latency caused by CPU or Disk bottlenecks, which includes the following scenarios:

• High latency driven by the CPU bottleneck, as discussed in ONTAP Data Processing latency - Resolution Guide. 
  • CPU resource is shared by all the aggregates on the same node, so CPU bottleneck can NOT be mitigated by volume migration to other aggregates on the same node. 
  • The busy volumes have to be migrated to other less utilized nodes. 
• High latency driven by the Disk bottleneck, as discussed in ONTAP Disk utilization and latency - Resolution Guide.
  • Disk bottleneck can be mitigated by migrating the busy volumes to other aggregates on the same node, or other aggregates from a different node. 
  • The target aggregates for the volume migration could be less utilized aggregates with the same type of disk drives, or they could be aggregates with faster disk drives. 
    • For example: Moving volumes from SATA aggregates to SSD aggregates would resolve most disk bottlenecks immediately.