Attila Craciun [:arny] Reporter
	Description • 6 years ago

PXE does not work, timeout, and if I want to select the HDD boot from one time boot it just get back in the menu.

	Mark Cornmesser [:markco] OOO 2024/04/15
	Comment 1 • 6 years ago

The node sis not seeing any bootable eufi devices. Open bug Bug 1467264.

	Mark Cornmesser [:markco] OOO 2024/04/15
	Comment 2 • 6 years ago

From Bug 1467264:

(In reply to Attila Craciun [:arny] from comment #3)
> Created attachment 8984031 [details]
> Screenshot from 2018-06-07 09-00-54.png
> 
> The server is slow while initializing and once is done, the attached error
> appear. 
> 
> 338-HPE RESTful API Error - Unable to communicate with iL0 FW. BIOS
> configuration resources may not be up-to-date. Action: Restore Manufacturing
> Defaults in HSU. 
> 
> I have tried to reset the BIOS to defaults but the same error.

Q: any thoughts?

	Q
	Comment 3 • 6 years ago

There are 3 separate firmware you can apply to the moonshot. Two of which hte MEz and Cartridge firmware might need updating to see the drive.  

This could also be a NAND issue here:
https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-c04996097

Before diving that deep is this one of the chassis the dhouse was updating ? It is possible it caused the Restful API to be blocked so the node could not register hence the 338 error.

	:dhouse
	Comment 4 • 6 years ago

This was reported last night in #ci also. I tested rebooting 281 to see the problem and it sticks at the pxe (looks like it never gets a connection out from the chassis).

I tried a reboot on cartridge #1 in the same chassis to see if it could be an issue for all of the chassis, but it boots correctly through pxe (gets dhcp, reaches the pxe/tftp and goes through normally).

I compared the configuration in bios and found no differences.

The chassis interface, and post info shows that the memory is all inserted (I had seen one cartridge not booting because the memory did not show as inserted).

I don't see any errors or anything strange during the boot. I also see it goes back to the boot after trying to boot from HDD. Could there be nothing on the disk to boot?

	:dhouse
	Comment 5 • 6 years ago

Mark, do you know if this host was working before, and do all the other windows servers in this chassis pxe boot? (may be a difference in dhcp between the linux and windows. since cartridge #1 pxeboots for me) This is on moon-chassis-7, and so I'm also wondering if there could be something odd in the switch configuration also since moon-chassis-7 was configured differently for testing lacp and xen.

	Mark Cornmesser [:markco] OOO 2024/04/15
	Comment 6 • 6 years ago

It was working previously. The other blades on this chassis are not exhibiting the same behavior.

	Attila Craciun [:arny] Reporter
	Comment 7 • 6 years ago

Reimaged, is back in action.

	Roland Mutter Michael (:rmutter)
	Comment 8 • 6 years ago

Looks good: https://tools.taskcluster.net/groups/IRiayNbAS1mmpo351Jb-cw/tasks/OVVm68hrQRatgHk1pqFpiA/runs/0 . Closing for now.