Dell 7010 Lenovo L530 with UEFI enabled, won't network boot.

[george1421]

@Sebastian-Roth Sorry to get you excited. I see I should have been clearer with my table.

UEFI pxe booting no == PXE booting in UEFI mode is not supported.

The only system that failed to boot via ipxe.efi is the 7010. All others that supported UEFI will boot the debug version either via USB or PXE (where available)

So the question I have now is, what is physically different between a o790 and o7010 ??? I’m finding it less likely that its a nic issue since both the 790 990 and 7010 use the same intel nic.

[george1421]

Continued testing:

e6410: UEFI not available.
e6420: UEFI available, UEFI pxe booting not allowed. Boot via efi flash drive using the debug ipxe.efi no problem.
e6430: UEFI available, UEFI pxe booting allowed. Booting either via PXE or generated debug messages (same as 7010. Also noted the 6430 and 7010 where purchased the same year).
e7440: UEFI available, UEFI pxe booting allowed. Booted correctly with either PXE or USB
e7240: UEFI available, UEFI pxe booting allowed. Booted correctly with either PXE or USB

OK, so I started digging into the details of the 7010 and the 6340 then comparing them to the 790 and 6420. The thing (from the spec sheet) that is in common between the 7010/6430 is the chip set with Q77/QM77 respectively. The 790/6420 has the Q65/QM67. I thought the graphics were the same between the two but on the 7010 we have the Intel HD4000 (integrated) and on the 6430 we have Nvidia Nvs 5200. All four systems use the intel 82579LM ethernet nic.

[Wayne Workman]

@george1421 said:

I’m finding it less likely that its a nic issue since both the 790 990 and 7010 use the same intel nic.

That is what I find most interesting…

[george1421]

Looking at the Lenovo L530 specs. That also uses the Intel HD 4000 graphics (same as 7010) and possibly the e6430 (I didn’t confirm that the exact computer I tested had the intel HD 4000 or the Nvidia) Both were options for the 6430.

The L530 chipset is the ivy bridge. The system was manufactured by Lenovo in 2012, which is about the same vintage as the 7010 and e6430 from my fleet. I don’t have enough details yet to draw a positive conclusion but there is a lot of similarities between the L530, o7010 and e6430

[Wayne Workman]

@george1421 So at work I have the worst possible set of models for iPXE with UEFI. Nice.

[george1421]

@Wayne-Workman Me too. I’m showing about 400 e6430s and ~800 7010s in our company.

We were going to start Win10 image development. Just after the first of the year. We may have to rethink our deployment tool if the ipxe wizard’s can’t figure this out. I’m a bit spurised that this issue hasn’t came up before. These 7010/e6430/L530s are 3 years old. If we play our cards right, we may age the Dells out before we get around to deploying Win10 to them.

[Wayne Workman]

@george1421 You can enable UEFI booting on the 7010 with “Legacy Option ROMs” turned on. It will use undionly.kkpxe for network booting just fine, and then boot to Win10 in UEFI mode just fine as well. You just have to rely on WOL for imaging I think, because once it’s booted in Legacy mode over the network, it can’t pass control to the HDD that wants UEFI.

It does work, you will not end the universe if you “cross the streams”. It’s just finicky.

[george1421]

@Wayne-Workman Interesting. (I’m still learning about UEFI) I thought to install a UEFI image, you had to boot in UEFI mode and place the image onto the system in UEFI mode otherwise the bios won’t be updated correctly. If we can boot in legacy mode and end up with a fully functional system booting in UEFI mode after deployment then this issue is not as big of a problem as it would appear.

[Wayne Workman]

@george1421 said:

If we can boot in legacy mode and end up with a fully functional system booting in UEFI mode after deployment then this issue is not as big of a problem as it would appear.

I’ve seen it done at work with Win8.1 and Optiplex 9020s, using undionly.kkpxe. The image that was used was UEFI, just Legacy option roms need enabled, and booting to the NIC first was not set in this particular instance - which is why I say rely on WOL.

[Sebastian Roth]

@Wayne-Workman @george1421 Are you still keen to try and get those in UEFI mode working? I’ve lost track of this but got a reply to this in the iPXE forum just lately. Nothing new yet but we might want to start digging again if you like? Possibly with some support in the iPXE forum as well.

If so I would ask to focus on one NIC/model at a time so we don’t get lost. The farthest we got is on the Dell 7010 (Intel 82579LM NIC) I think. I’d go into this - as well because iPXE detects the same NIC on a different PC I tested (Fujitsu ESPRIMO P910 E85+) without any issue. On the other hand this might be because of an UEFI firmware bug on the Dell machine which would take us a long way to get it fixed. But we don’t know yet.

But we can also go for a different machine/NIC if you like. I am open for everthing. Just let me know. Maybe just open a complete new thread and link this discussion here as a reference. I am more than happy to provide information about how to compile iPXE binaries so you don’t depend on me for that!

[george1421]

I’ve been kind of busy getting big project at work to the finish line, but I’ve been thinking about attempting to use (as a test) other boot loaders like grub or elilo to see if they handle uefi booting better than ipxe. But right now its not clear if we have 2 or 3 different issues. Is it the built in uefi bios, the boot loader or the current FOS kernel that is having problems. It may be a combination of all three (which makes it even harder to debug).

At this point we need to eliminate 2 of the three possibilities to work on the third.

[george1421]

This is an interesting article for efi booting a vm client. https://communities.vmware.com/docs/DOC-28494

Interesting points are efi requires 96MB of ram where bios fits in 4MB of ram.

There were a couple references to this “with a partition scheme suitable for EFI”. Which makes me wonder if there has to be something done special with the init fs to make it efi suitable?? Because for me where it is crashing is mounting the rootfs. The kernel boots, but I’m getting an invalid or corrupt initfs.

This is more rambling talk right now than fact.

[Sebastian Roth]

@george1421 said:

There were a couple references to this “with a partition scheme suitable for EFI”. Which makes me wonder if there has to be something done special with the init fs to make it efi suitable??

Interesting. Haven’t thought about this before. Well we do have FOG (FOS) running on several different UEFI machines. So I am certain that it works with those inits Tom is building. AFAIK there is no partition scheme involved. initrd is just a compressed image file - no MBR, no partitions, no GPT …