ONTAP SAN Protocol Latency VMware Connected via Blade Chassis Interconnects
Applies to
- NetApp AFF
- ONTAP 9
- VMware ESXi 7.0 Update 3g
- HPE Blade Chassis with Ethernet/FCoE interconnects
- Brocade Fibre Channel Switches
Issue
- VMware hosts and applications experienced high protocol latency, despite low storage volume latency on the NetApp storage.
- Latency spikes were observed in both VMware and ONTAP protocol metrics, particularly affecting production VMs (e.g., Kafka cluster).Latency values exceeded 100ms, with peaks up to > 200ms .
- esxtop we continue to see high DAVG values to collect esxtop batch data at time of issue
WARNING:ScsiDeviceIO:1513:Device naa.600a0980383 performance has deteriorated. I/O latency increased from average value of 20762 microseconds to 419996 microsecondsesxtop
DEVICE DQLEN WQLEN ACTV QUED %USD LOAD CMDS/s READS/s WRITES/s MBREAD/s MBWRTN/s DAVG/cmd KAVG/cmd GAVG/cmd QAVG/cmd
naa.600a0980383 64 - 53 0 0 0.00 256.66 62.58 194.07 2.11 92.62 99.21 15.48 114.69 15.47
naa.600a0980383 64 - 49 0 0 0.00 179.29 41.60 137.69 3.60 61.30 186.69 87.77 274.46 87.75
naa.600a0980383 64 - 64 108 0 0.00 1.48 0.20 1.29 0.01 0.62 226.75 266.06 492.81 266.04