SAN Boot Issues

[syedaliferoz.sm.ap1]

We have a Server configured with SAN Boot, where we zoned and configured the target details. But we had rebooted the server for some request after which server is not booting to OS.

Zoning is configured Properly.

Boot Policy Configured Properly

Please help on the same.

[marce1000]

         - What is the server model ?
         - Post the boot process (as far as it goes).

 M.

[syedaliferoz.sm.ap1]

 Server Model : UCS B200 M4 

Boot process :  We installed the OS via PXE boot and installed the OS on Boot Lun Configured and delivered already. but after recent reboot, boot lun is not detecting in the KVM Console and got struck in PXE message .

We verified the HBA's VSAN's Boot Policy and LUN ID which are all looks good that made connectivity good.

Also, this is happening after we swap the HW from B480 M5 to B200 M4 and upgraded to Oracle Linux Release 8.x with Redhat Kernal.

[Kirk J]

Has the B200M4 ever booted correctly to SAN?

Did you just try to re-apply the service profile previously on the B480 M5 to the B200M4?  Or did you reinstall OS after different hardware used with service profile?

There are likely different VIC cards in use, which may/may not have an impact.

Do you see the service profile's WWPNs flogi'd in? (connect nxos, show npv flogi-table ) 

Kirk...

[Steven Tardy]

Are you using Legacy BIOS boot mode or UEFI boot mode?

If you are using UEFI mode then the OS is responsible for adding a BIOS NVRAM boot entry. If BIOS was cleared or the kernel upgrade did not properly update the BIOS NVRAM, then the server won't boot the OS and fail back to PXE like you are seeing.

Adding the BIOS NVRAM option through the BIOS is usually the quickest way to boot the server:

1. F6 (or whatever the key is to get into BIOS configuration.)
2. Boot Options
3. Add boot option
4. Path for boot option
5. drill down to the file system EFI\Boot directory and select BOOTX64.EFI (or whatever redhat is calling their boot shim).
6. Save and exit.

UCSM has the ability to configure the BIOS NVRAM entry, but it is often more simple to let the OS do it (until someone does a [clear CMOS] removing the UEFI entry).

[WUR]

Did you open the SR with TAC.

Boot lun is not detecting in KVM console and gor stuck with PXE message ? >>> What is the message?
How did you verify the HBA's VSAN's Boot Policy and LUN ID?
Did you follow proper procedure for Hardware downgrade from B480 M5 to B200 M4? What is the reason for downgrading the hardware?
Did you check the compatibility before upgrading to Oracle Linux Release 8.x with Redhat Kernal.?

[syedaliferoz.sm.ap1]

We had rebuilt the server and resolved the issue.

Thank you all for your inputs