ESXi 6.5: How do I increase the "No of outstanding IOs with competing worlds:" greater than 64

[fdrietatns]

I'm trying to mach the lun qdepth with the HBA qdepth and am getting a hard error that i can't set it beyond '64' [below]:

Unable to set device's sched-num-req-outstanding. Error was:Cannot set device queue depth parameter. sched-num-req-outstanding should be <= 64

I haven't found anything yet regarding this so i'm hoping someone can help me remove this artificial error. I am able to freely set this option on 6.0 without incident.

Thanks.

[fdrietatns]

This article suggests i ought be able to set it between 1 and 256 :VMware Knowledge Base

vSphere 5.5, vSphere 6.0 and vSphere 6.5

To set the VMkernel limit for a particular LUN/device in vSphere 5.5 and 6.0 using the command line:

1. Open an SSH session to the host. For more information, see Using Tech Support Mode in ESXi 4.1 and ESXi 5.x (1017910).
2. To check the current value for a device,run the command:
   
   esxcli storage core device list -d naa.xxx
   
   Note: The value appears under No of outstanding IOs with competing worlds:
   
3. To modify the current value for a device,run the command
   
   esxcli storage core device set -d naa.xxx -O Value
   
   Where Value is between 1 and 256 (vSphere 5.5 and 6.0) and between 1 and Max Queue depth of the device (vSphere 6.5)

I haven't rebooted the hosts since i changed the HBA qdepth so perhaps that will rectify this for me.

[fdrietatns]

Using "esxcli storage core set -m" has no effect. No error is generated but the value isn't applied.

Based on the language used in the KB, it looks like the -O option is constrained to the -M option but using -M has no effect. I have already confirmed that the qlfxmaxqdepth parameter is set correctly but am unable to effect change at the device level.

[fdrietatns · Accepted solution]

Disregard, I set the qlfx kernel driver and not the ql2x.

After setting the ql2x driver param and rebooting everything is working as expected.

Not sure how i missed that entirely.