What is the difference between SVM latency and volume latency in AIQUM?
Applies to
- Active IQ Unified Manager (AIQUM)
- ONTAP 9
Answer
- SVM latency is displayed by selecting Performance: All Storage VMs view in STORAGE > Storage VMs. This is based on the IOPS and time to respond to the requests per SVM.
- Volume latency is displayed by selecting Performance: All Volumes view in STORAGE > Volumes. This is based on the IOPS and time to respond to the requests per volume.
- SVM latency does not mean the average latency of all volumes under the SVM.
Example:
Assuming the SVM contains two volumes, volumeA and volumeB, with following statistics:
volumeA : IOPS 100[op] and Time to respond 1000[ms]
volumeB : IOPS 400[op] and Time to respond 12000[ms]
- The volume latency of volumeA is 10[ms/op] calculated by
1000[ms]/100[op]
- The volume latency of volumeB is 30[ms/op] calculated by
12000[ms]/400[op]
- The SVM latency is 26[ms/op] calculated by
(1000+12000)[ms]/(100+400)[op]
- The volume latency of volumeA is 10[ms/op] calculated by
Note: AIQUM gethers these performance statistics from ONTAP cluster every 5 minutes