MSA SAS SAN with VMWare vSphere /ESXi with Multi pathing

[mr_vaughn]

Been using HPE MSA SAN since the P2000 days. But in all the guides over the years I have not seen a specfic case and use of SAS over the Fiber Channel/iSCSI deployments. 

I tend to deploy 2 DisK pools, each with different RAID set following the power rule. and have read cache's on each. Reason for 2 Disk Pools are the use on the both  Controllers cache which can help quite a bit of performance.

Is it the same procedure for all FC/iSCSI & SAS on the below?

1. Multipathing
https://www.hpe.com/psnow/doc/4aa4-7060enw

is missing the commands. Found these for 6.x Do they still work with 7.x /8.x

esxcli storage nmp satp set -s VMW_SATP_DEFAULT_AA -P VMW_PSP_RR
esxcli storage nmp satp set -s VMW_SATP_ALUA -P VMW_PSP_RR

An older all in 1 command

for i in 'esxcfg-scsidevs -c |awk '{print $1}' | grep naa.600c'; do esxcli storage nmp psp roundrobin deviceconfig set --type=iops --iops=1 --device=$i; done

2. Some say  to Adjusting Round Robin IOPS limit from default 1000 to 1  but that seem to be the old 1Gb ISCSI days
page 10/21 in the "Deployment Considerantions" states "Best practice: Keep the default IOPS and bytes limit. HPE MSA storage arrays do not see a benefit by changing these limits." 

https://knowledge.broadcom.com/external/article?legacyId=2069356

[ArunKKR]

Hi,

For any version (FC/iscsi/SAS) of MSA G6 arrays the recommended path selection policy is round robin.

As you have correctly identified HPE MSA 2060 storage arrays do not see a benefit by changing the IOPS limits. 

For older model arrays G5 and G4 you may still follow the IOPS setting of 1.

https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.storage.doc/GUID-AE95818C-0782-429B-9526-DF12673A63C7.html

NOTE: The above-mentioned URL will take you to a non-HPE website. HPE does not control and is not responsible for information outside the HPE website.

Below command could be used to identify naa id of MSA volumes and current path selection policy setting.

 esxcli storage nmp device list

Execute below command to change PSP to round robin:

esxcli storage nmp device set  --device naa.600c0ff000533a9fb58fd56601000000 --psp  VMW_PSP_RR

Execute below command to change IOPS value to 1:

esxcli storage nmp psp roundrobin deviceconfig set --type "iops" --iops 1 --device naa.600c0ff000533a9fb58fd56601000000

MSA volumes usually have an naa id starting with 600c.

The naa id for each volume will vary

[ArunKKR]

Hi Vaughn,

 
Let us know if you were able to resolve the issue.
 
If you have no further query and you are satisfied with the answer then kindly mark the topic as Solved so that it is helpful for all community members.

[mr_vaughn]

There was a command also to also set the default path selection to Round Robin for MSA volumes.

So on creation of new LUN's /VMFS Datastores it will default fo Round Robin.

[ArunKKR]

Hi,

I guess you are referring to vmware SATP rule.

These are the sample commands to create SATP rules for MSA 2040 and 2050.

esxcli storage nmp satp rule add -P=VMW_PSP_RR -s=VMW_SATP_ALUA -O "iops=1"-V=HP -c="tpgs_on" -e="HPE MSA 2040 Custom Rule"


 esxcli storage nmp satp rule add -P=VMW_PSP_RR -s=VMW_SATP_ALUA -O "iops=1"-V=HPE -c="tpgs_on" -e="HPE MSA 2050 Custom Rule

Please note that my expertise lies with MSA storage. You may wait for any inputs from vmware experts to check whether the above commands would work and the correct commands for all versions of ESXi.

https://docs.vmware.com/en/VMware-vSphere/7.0/com.vmware.vsphere.storage.doc/GUID-D10F7E66-9DF1-4CB7-AAE8-6F3F1F450B42.html

NOTE: The above-mentioned URL will take you to a non-HPE Web site. HPE does not control and is not responsible for information outside of the HPE Web site

[mr_vaughn]

@ArunKKR 

Here is an older commands that used to work that I used.

esxcli storage nmp satp set -s VMW_SATP_DEFAULT_AA -P VMW_PSP_RR
esxcli storage nmp satp set -s VMW_SATP_ALUA -P VMW_PSP_RR

for i in `esxcfg-scsidevs -c |awk '{print $1}' | grep naa.600c`; do esxcli storage nmp psp roundrobin deviceconfig set --type=iops --iops=1 --device=$i; done

but the last one does not seem to work any more