If you have recently referenced the FNIC tunables doc at https://www.cisco.com/c/dam/en/us/solutions/collateral/data-center-virtualization/unified-computing/guide-c07-730811.pdf and try to adjust the max qdepth parameter in ESXi, you will get a syntax error from ESXi.
# esxcli system module parameters set –p fnic_max_qdepth=128 –m fnic
should actually be
# esxcli system module parameters set –m fnic –p fnic_max_qdepth=128
The module needs to be referenced before the parameter.
There is an open request to have this corrected in the doc, but wanted to call this out as I have had several cases recently where this has caused issues for customers needing to change the fnic qdepth from its default setting of 32.
Your San storage vendor should be the one to set the general guidelines for qdepth.
If you are running some very high storage I/O apps like MS SQL or Oracle, then you may need to change the IO throttle (in FC adapter policy in UCSM) and Lun Qdepth from their defaults (usually this is being changed to higher values, i.e. from default of 32 to 64) but you need to check with your storage vendor's recommendations for the specific array you have.
Thanks,
Kirk...
