Write latency and low throughput on Metrocluster Environment
Applies to
- ONTAP 9
- Metrocluster
Issue
- High write latency on Metrocluster systems
- Low number of IOPS can be seen than the expected IOPS
- From CLI, Check using command
qos statistics volume latency show
Example:
cluster::> qos statistics volume latency show -node node1
Workload ID Latency Network Cluster Data Disk QoS NVRAM Cloud
------------ ------ ---------- ---------- ---------- ---------- ---------- ---------- ---------- ---------- -
volume_1 12578 10ms 1ms 0ms 0.5us 1ms 0ms
9ms 0ms
volume_2 5232 2ms 1ms 0ms 0.25us 1ms 0ms 0ms 0ms
volume_3 17160 1ms 1ms 0ms 0.25us 1ms 0ms 0ms 0ms
The above example means that the volume_1
is impacted by NVLog Transfer latency