modify the storage qdepth for only some few Devices,

[momox1]

Hello Guys,

I'm working on a script to increase the storage device qdepth in only some selected LUNs(full flash), but I'm a bit stuck on how  can I match the datatore name  containing "flash" string with the naa.

The idea is to target several vcenters.

****************

esxhost=get-vmhost

foreach ($esxhostlist in $esxhost)

{

$esxcli = Get-EsxCli -VMHost $esxhostlist.name

#this command is parsing all the devices.
$list=$esxCli.Storage.Core.device.list()|where {$_.IsLocal -eq "false" -and $_.DisplayName -match "Hitachi"}|select device,DeviceMaxQueueDepth    ##how can I target only "Platinum" datastore???

write-host "Setting qdepth=16 to $($esxhostlist.name)"
$list| foreach {$esxcli.storage.core.device.set($null,$null,$_.device,$null,$null,$null,4,$null,$null,$null,$null,$null,$null,$null) |Out-Null}

}

************

I appreciate your help.

[LucD]

Can you try like this

$dsName = 'Platinum'

Get-VMHost -PipelineVariable esx |
ForEach-Object -Process {
    $esxcli = Get-EsxCli -VMHost $esx -v2
    $esxcli.storage.vmfs.extent.list.Invoke() | where{$_.VolumeName -eq $dsName}  |
    ForEach-Object -Process {
        $esxcli.storage.core.device.set.Invoke(@{device = $_.DeviceName; maxqueuedepth = 4 })
    }
}

[momox1]

 I get the following when I try

Message: EsxCLI.CLIFault.summary;
InnerText: Unable to find device with name EsxCLI.CLIFault.summary
At line:6 char:9
+ $esxcli.storage.core.device.set.Invoke(@{device = $_.DeviceNa ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : OperationStopped: (:) [], MethodFault
+ FullyQualifiedErrorId : VMware.VimAutomation.Sdk.Types.V1.ErrorHandling.VimException.MethodFault

[LucD]

The pipeline symbol was missing, I corrected the code above

[momox1]

Hello LucD,

Thank you for the answer. The script is working .I get this when I  try the put  the max qdepth per device at 16 : 

Message: EsxCLI.CLIFault.summary;
InnerText: Unable to set device's max-queue-depth or current-queue-depth. Error was: Unable to complete Sysinfo operation. Please see the VMkernel log file for more details.: Bad parameterEsxCLI.CLIFault.summary
At line:6 char:9
+ $esxcli.storage.core.device.set.Invoke(@{device = $_.DeviceNa ...
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ CategoryInfo : OperationStopped: (:) [], MethodFault
+ FullyQualifiedErrorId : VMware.VimAutomation.Sdk.Types.V1.ErrorHandling.VimException.MethodFault

********************* When I Look into the logs :

2021-04-28T09:19:04.861Z cpu29:2100215 opID=c74a7459)ScsiVsi: 3731: Can't set the maxPathQueueDepth value to more than device advertised maxPathQueueDepth 4

[momox1]

Completing the information. I'm checking why isn't accepting 16 as the current hba's

 esxcli system module parameters list -m nfnic
Name Type Value Description
------------------------ ----- ----- -----------
ecpu_ka_timeout ulong nfnic ecpu keep alive timeout: Default = 10. Range [10 - 120] seconds. 0 to turn off.
log_throttle_count ulong nfnic log throttle count: Default = 64
lun_queue_depth_per_path ulong 16 nfnic lun queue depth per path: Default = 32. Range [1 - 1024]

[LucD]

That is afaik a HW limitation, can't really change that with a script

[paudieo]

looks to be driver related , you prob have to bump up lun_queue_depth_per_path

Check if your hba/driver supports this

https://www.cisco.com/c/en/us/support/docs/servers-unified-computing/vmware-esxi-cisco-ucs/214808-co... 

https://communities.vmware.com/t5/VMware-PowerCLI-Discussions/Unable-to-nfnic-queue-depth/td-p/22966...

[momox1]

Yes, thank you for the links, we have to reboot some of the hosts so the hba can apply the right qdepth.

I'm wondering if there is a parameter of module where I can find the configured qdepth vs the current applied...

Thank you for your help.