Increased read latency on a node due to increased FabricPool cold tier reads
Applies to
- ONTAP 9
- FabricPool
Issue
- Latency has suddenly increased on a node after a specific date
- This can be seen in both Active IQ Unified Manager and Active IQ
- Example: latency was 1 ms on September 23, but September 24 is 25 ms or more for reads
- Multiple volumes report "Latency Critical Threshold Breached" alerts.