iLO 5 v3.0 firmware does not correctly see disks via SNMP GET, GET-NEXT and WALK

[Michael_Adm]

I have two Gen10 DL380 servers with almost the same configuration.

1-server:

2 x Intel(R) Xeon(R) Gold 6242R

1 x HPE Smart Array P408i-a SR Gen10 (connected to one "8Sff Sas/Sata Bc Box" dirrectly with 8 SAS-SSD disks)

1 x HPE MR416i-p Gen10+ (connected to one "8 SFF 16G x4NVMe UBM4 BC BP" dirrectly with 4 NVMe disks)

2-server:

2 x Intel(R) Xeon(R) Gold 6242R

1 x HPE Smart Array P408i-a SR Gen10 (connected to one "8Sff Sas/Sata Bc Box" dirrectly with 8 SAS-SSD disks)

2 x HPE MR416i-p Gen10+ (connected to one "8 SFF 16G x4NVMe UBM4 BC BP" dirrectly with 8 NVMe disks)

1-server SNMP Walk:

27.01.2024 12:08:31 (1 ms) : Device: 172.27.0.45
27.01.2024 12:08:31 (2 ms) : SNMP V1
27.01.2024 12:08:31 (3 ms) : Walk 1.3.6.1.4.1.232.3.2.5.1.1.3
27.01.2024 12:08:31 (6 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.1 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (8 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.2 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (10 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.3 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (13 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.4 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (15 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.5 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (17 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.6 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (19 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.7 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (22 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.8 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:08:31 (25 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.13 = "INTELSSDPF2KX076TZ" [ASN_OCTET_STR]
27.01.2024 12:08:31 (28 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.14 = "INTELSSDPF2KX076TZ" [ASN_OCTET_STR]
27.01.2024 12:08:31 (31 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.15 = "INTELSSDPF2KX076TZ" [ASN_OCTET_STR]
27.01.2024 12:08:31 (34 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.16 = "INTELSSDPF2KX076TZ" [ASN_OCTET_STR]

2-server SNMP Walk:

27.01.2024 12:12:39 (5 ms) : Device: 172.27.0.46
27.01.2024 12:12:39 (7 ms) : SNMP V1
27.01.2024 12:12:39 (9 ms) : Walk 1.3.6.1.4.1.232.3.2.5.1.1.3
27.01.2024 12:12:39 (14 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.1 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (20 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.2 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (23 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.3 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (27 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.4 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (31 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.5 = "HBTAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:12:39 (34 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.6 = "HBCAC2DH2SUN3.2T" [ASN_OCTET_STR]
27.01.2024 12:12:39 (38 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.7 = "INTELSSDPF2KX038TZO" [ASN_OCTET_STR]
27.01.2024 12:12:39 (43 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.8 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (47 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.11 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]
27.01.2024 12:12:39 (51 ms) : 1.3.6.1.4.1.232.3.2.5.1.1.3.0.14 = "INTELSSDPF2KX038TZ" [ASN_OCTET_STR]

In the WEB interface of iLO 5 in "System Information - Storage" all controllers and disks are displayed correctly on both the first server and the second server (second server have 8 sas-ssd and 8 nvme disks).

Of course, I understand that this is not a standard configuration, but everything works perfectly, except for this unpleasant moment with SNMP on the second server that has two NVMe controllers.

I do not have a paid HPE service and cannot transfer this information to them directly. Therefore, I ask for help from those people who can do this.
Thank you in advance.

[support_s]

System recommended content:

1. HPE ProLiant Gen10 Servers-(iLO 5) - How to fix Incorrect Time Stamp on iLO Event Log Entries

2. HPE ProLiant DL380 Gen10 Server - Blank SMH Screen

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

Thank you for being a HPE valuable community member.

[Michael_Adm]

I don’t quite understand what the SNTP protocol has to do with it (it is configured correctly on the servers and the time in iLO is synchronized without errors).
In relation to AMS, the latest version is installed - ams.exe 2.51.4.1 agentless management service.

[Tam92]

Try to reset iLO and restart the AMS service once

[Michael_Adm]

What I tried: rebooted the OS, restarted iLO, even reinstalled the OS again with all new drivers.
Did not help.
But when I turn off one of the two HPE MR416i-p Gen10+ in the BIOS, SNMP WALK becomes normal - it displays all 8 SAS-SSDs and four NVMe disks, similar to the first server.
I turn on the second one (with a restart of the OS, of course) in the BIOS and everything returns as described in the title - it displays eight NVMe drives and two SAS-SSDs.
It seems that NVMe overwrites SNMP information about the SAS-SSD.

[Tam92]

Hello,

Try to put the other MR controller in a different slot controller by Processor 2 and check.

https://support.hpe.com/hpesc/public/docDisplay?docLocale=en_US&docId=a00008180enw

Thanks,
TAM

[Michael_Adm]

I tried - no changes, that both MR controllers are on the first or second CPU, that they are spaced apart, one on the first CPU, the second on the second CPU - the same way.
I’m almost sure that this is a matter of the logic of the iLO 5 program.
The iLO programmers did not expect that the server could have two MR controllers and a third SR controller.

[Michael_Adm]

@Tam92 

Updated to the latest firmware iLO 5 v3.01. Without changes.
SNMP works the same way as in the previous version of iLO 5 - NVMe drives overwrite SNMP information on the SAS drives of the SR controller.