failing boot disk on NS204i-u boot controller issues

[dmateman]

My HP server didn’t report that a disk had failed on the boot controller. I have two NVMe drives in the boot controller. The server wouldn’t boot anymore, but it also didn’t show any lights indicating something was wrong. Still, it wouldn’t boot. After removing the faulty NVMe, it booted again and then it did indicate that one disk had failed, and the LEDs started showing that something was wrong.

Is this normal, or does anyone have experience with this? I mean, such an expensive HP server should clearly show in the logs what’s wrong and which disk it is, instead of only saying that PCI port 7 (NS204i-u) is not reachable.

And the seccond question is

The server suddenly went offline, and the boot controller is running in a RAID 1 configuration, so if one disk fails it should just keep running, right? So why does it only work once I remove the defective disk? It seems similar to RAM: if one RAM module is faulty, it won’t boot.

[Mamatha_J]

Hi @dmateman 
Please take a look at the documents, you might find them useful. 
The official HPE documentation covering the behavior, configuration, and installation of the NS204i boot device (which includes the HW RAID1 logic for the two NVMe SSDs).
HPE NS204i Boot Device User Guide: https://support.hpe.com/hpesc/public/docDisplay?docId=sd00001830en_us&docLocale=en_US&page=GUID-ED109CA6-8C7B-4CAF-9F8C-B9B3E78235AA.html&utm_source
HPE Server Manuals – NVMe/LED Status Definitions:HPE ProLiant DL380 Gen11 Server User Guide — NS204i-u LED Info: https://support.hpe.com/hpesc/public/docDisplay?docId=sd00002446en_us&page=GUID-3325A88D-E92C-4CA4-AA47-52383465F004.html&utm_source
HPE ProLiant DL360 Gen11 Server / NS204i-u LED: https://support.hpe.com/hpesc/public/docDisplay?docId=sd00002445en_us&docLocale=en_US&page=GUID-3325A88D-E92C-4CA4-AA47-52383465F004.html&utm_source
HPE NS204i-u Boot Optimized Storage Device Data Sheet: https://www.hpe.com/psnow/doc/PSN1014691212ILEN?utm_source

Thanks & Regards,

Mamatha Ajaraddi

[shiva_jr]

Hi @dmateman , 
  Yes, you are right. As per my understanding-

a. I assume, the NS204i‑u cannot report the actual state of each NVMe drive until the it initializes again after POST. Because of this, you won't get notified even a disk failed.

b. The boot device enables the deployed OS to be mirrored through a dedicated hardware RAID 1.
   As per the document. the NS204i-u works auto creates RAID 1 volume during boot. t

What should be done now? 
1. Identify the failed disk, replace it.
2. Reboot and verify whether it creates a RAID 1 with fresh initialization.

Regards,
Shiva_jr