Slowness on NFS clients with little or no latency from ONTAP
Applies to
ONTAP 9
Issue
- High latency or a sudden drop in throughput during peak load on NFS clients would be observed.
- Oracle DB reports high wait time on DB commits.
- It cannot be attributed to a CPU/Memory/Disk bottleneck on the storage controllers via CLI or monitoring tools.
- From ONTAP 9.8, Event Log error messages similar to the following:
Fri Jan 01 01:47:19 +0100 [NETAPP03: kernel: nblade.execsOverLimit:error]: The number of in-flight requests from client with source IP 10.20.30.40 to destination LIF 10.20.30.50 (Vserver 7) is greater than the maximum number of in-flight requests allowed (128). The client might see degraded performance due to request throttling.