HPE DL360 Gen11 with NVME Raid 5?

[techdmc]

Hi, I've used HPE DL360 Gen10 Server with a P408i-p SR Gen10 Controller with SATA SSDs in the past. NVMe weren't supported in a Hardware RAID 5 Configuration with this hpe server.

Now I've read about Gen11 Servers in combination with HP MR408i-o Gen 11 Controller. In Specs, the MR408i-o Gen11 supports PCIE NVME Devices, but i couldnt find in the docs, if it is really possible or only in HBA mode.

Does there any document exist for that?

At the end, I want to build a RAID 5 with 6x PCIE NVME SSDs.

[Cali]

Here is the document: 400 series | HPE MR Gen11 Controller User Guide

And yes, it supports Raid 5 with NVMe Disks.

(MR Controller support R5 and R6 only with Battery Cache)

Cali

[techdmc]

Thanks for your reply @Cali 

What storage controller is exactly needed to use 6x PCIE 4.0 NVMe in a RAID 5? I've read, that some controllers only support up to 4 NVME drives.

[Cali]

The 408 supports 8 NVMe Drives and the 416 16 NVMe Drives.

You may also need the appropriate hard drive cage and cable.
The best thing would be to take the configuration to an HPE partner, they have a tool that checks whether the configuration works as it is or whether something else is needed.

Otherwise, you would have to put it together using the QuickSpecs.

https://www.hpe.com/psnow/doc/a50004307enw.pdf?jumpid=in_pdp-psnow-qs

Cali

[support_s]

Hello,

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.

Please click on "Thumbs Up/Kudo" icon to give a "Kudo".

Thank you for being a HPE valuable community member.

[Manuel Boosch]

Hello,

i know this thread is old but i had issues with using the f.e. MR408i-o Gen11 Controller in a DL380 Gen11 in vmware ESXi's passtrough mode (directly passtrough the mr-controller into a virtual machine) in combination with 8xNVME's attached to it in Raid5. If we tried to write data to the Raid5 (8x1,6TB NVME) data was always corrupted. Firmware & Drivers were all the latest from Jan 2025. If we removed the passtrough mode from the mr-controller and recreated the array using it as Raid5 local-vmfs-datastore the problem went away. 

While troubleshooting we found an 3 year old HPE-KB which described this situation very clearly:

https://support.hpe.com/hpesc/public/docDisplay?docLocale=en_US&docId=a00128391en_us 

After adding the line pciHole.start = "2048" to the vmx file of the virtual machine using the mr-controller in esxi passtrough mode all problems with corrupted data were gone... and never came back

This KB should be updated. Because our controller experienced the same issues and is not listed. And this problem is happening with all MR-Controllers in Gen11, too.

But it only applies if all these conditions are met:

- using HPE MR-Controller

- only applies if the MR-Controller has NVME Drives attached

- using the MR-Controller in VMWare ESXi's passtrough mode into a vm

- the vm has more then 2 GB of RAM

So in general RAID5 on nvme Drives with MR-Controller is supported, yes but if paastrough mode in vmware is used - then vmx needs adjustment (regardless which os - nearly every vm needs more then 2 GB ram if a controller is attached to it)

Regards

PS:

There is another HPE-KB describing the issue:

https://support.hpe.com/hpesc/public/docDisplay?docId=a00131053en_us&docLocale=en_US