Begging for your Help - HP Micro G8 / B120i / LSI MegaRAID / HexOS - HELL

[Sam of Ham]

Hi all,

I'm truly sorry to be bringing you all a barely-HexOS related problem, but I come begging for help from the people who likely know best--NAS owners, Self-Host Enthusiasts, and the HexOS team themselves--with my proverbial heart in my hands. I'm at the point where I am truly frustrated and I just need to ask in case anyone has the magic solution for me (and I will owe you big time if you do!)

Mods, I more than understand if this isn't relevant and you're welcome to remove this, and I apologise for wasting anyone's time. 

Here's the story. I'm upgrading from a HP Microserver (circa 2010) with no special upgrades; mainboard, SATA bays, TrueNAS CORE. Over time, the TrueNAS system locks up, and, while it works, I can't access or log into it at all - the login page just gives me "incorrect username/password" and I have to force power-cycle the thing. That's fine; I've wanted to upgrade for a while and found a good deal on a fully-kitted HPE Microserver G8 (with iLO - perfect!) from an enthusiast. HBA card, integrated RAID controller, 4-bay SATA, iLO, Xeon CPU, all the mods to make Disk 5 work for bootable media.

I have had nothing but problems in migrating. Wanting to get off of old CORE (iX are discontinuing it, plus, doing moronic things like removing native SMART data...) to HexOS and SCALE, and since I have an EA backer license wasting away, I figured, no worries, I'll install HexOS and migrate the disks. I plugged in an SSD, installed HexOS, claimed the server, happy as. I backed up all the data a second time just in case and moved the disks across (4x4TB WD Reds, FWIW.) No disks in HexOS/SCALE interface. Alright... Probably RAID - it used to be software RAID handled by CORE, this thing has the B120i. That's fine. Go into HP SSA - no disks, only bay 5. WTF?

Skip a lot of time and headaching and I realise the backplane is connected to the PCIe RAID controller (LSI 9261-8i) not the board (should have realised that, I know. I'm new to HBA RAID beyond just SW raid on Windows and CORE.) I... can't work out how to use the RAID card, it doesn't use a GUI, so it's all magic unknown CLI nonsense (to me, at least). Also, with some painstaking searching - it looks like LSI RAID cards need to be put into HBA mode for SCALE to control the RAID, but, SCALE doesn't support the card in HBA mode. Okay - fine. F*ck the RAID card - I disable it in BIOS and go back to the B120i. At this point I'm content with the RAID controller potentially losing me my data during RAID config. I screw around with boot modes about 18 times and finally get the drives to show up in HP SSA. Hoo-freaking-ray. Set up RAID, off we go to boot HexOS and install and get my goddamn NAS back up and running. 

And now we're up to NOW. With the drives in, I get the below: 

And with the drives OUT, I get this!

So I can't use the RAID controller, I maybe can't use the B120i, and it almost feels like I can't use HexOS either. Argh! I am lost! I feel like I'm destined not to have a nas with any form of modern features! No iLO, no RAID controller, no damn apps, just my buggy old CORE interface on a nearing-20 year old box. 

So, for at least just the sake of your understanding of my current mindset, I plead for any help anyone might have for me. I wrote this a little tongue-in-cheek-humoured, but honestly, I'm so damn angry at this whole process (not HexOS, as I said, but the whole thing; HP, SSA, the B120i, the LSI card, the iLO, the NAS, CORE, SCALE, HexOS, you name it, I'm just soured on the whole lot, meanwhile my NAS is down and my hopes of a simple home freaking media server are shrivelling up.) 

Thank you for your time and for any help you could possibly offer me. 

[Sam of Ham]

OH, by the way - yes, I did re-create the HexOS bootable drive, in case something on read screwed up GRUB or the HexOS Image. I used Rufus 4.6 and the lates HexOS .iso. 

[Manders]

Hey there!

Can you clarify something, because that text was a lot 😄

1. Have you got HexOS and TrueNAS Scale up and running now, albeit without any HDD pool?
2. Are you trying to use your existing 4x4tb HDD in HexOS
3. You mention RAID a whole lot, but hardware and software RAID. But HexOS and TrueNAS only support OpenZFS as far as I am aware, so is your existing pool in the OpenZFS format, or actually RAID?

[Sam of Ham]

Thanks @Manders!

1. I HAD HexOS running at the start but couldn't get any drives to work - now I can't get it to install anymore (see above screenshots.) 
2. Yep, though I don't mind if the data is wiped (would prefer not to but I know it may be necessary)
3. Current pool is a bit of a mystery but it was handled by CORE. ZFS it would be I believe as I have the warning to upgrade my ZFS flags in the CORE install. No hardware raid in the old system, but I COULD run hardware raid in the new NAS box. 

Thank you! Hope that helps!

[PsychoWards]

Hey, 

1. The drive you are installing on, does it already contain an OS/was it used before or is it a new drive? If it was used in the past, can you please try wiping? This helped some people in here already which were also not able to install Hexos on an used drive.

2. Try installing Hexos without any data drives connected and then login into the UI and use following guide to migrate from Core to Hexos, you should follow the path for migrating to 24.10 or later https://www.truenas.com/docs/scale/24.10/gettingstarted/migrate/migratingfromcore/

3. Never use HW raid in combination with Hexos/Truenas, let the OS handle everything in SW raid. HW Raid + SW Raid gives you the disadvantages of both but none of the advantages 🙂 

[Sam of Ham]

Hey @PsychoWards, thanks for your help. 

Good point - I HAD wiped the drive prior to installing HexOS the first time, BUT not during the most recent round of struggling through the LSI and B120i cards. I sort of assumed HexOS would still boot, and then when it didn't, figured I would just re-run the installer - I don't think I wiped the disk. I'll do that tonight and let you know how I go, thank you!

Screenshot 2 is currently what happens without drives connected - suspect the above might help though. I thought to recreate the HexOS image on the USB drive, but not to wipe the destination drive!

Got it, thanks on the RAID info. I had assumed that, if I let a dedicated RAID system handle all that, and just point TrueNAS towards the partitions, I'd be getting the benefits of both, but I can see why that would be the opposite of what I'm getting. If you'd like to, I'd love to learn more about why it's better to go with one or the other, especially when it comes to ZFS - my understanding of ZFS is basically "it's a file system and disk manager" (yes, I know that's like saying Windows is a file system... But that's what I know right now!!)

Thanks heaps. Will report back tonight. 

[PsychoWards]

Hey,

please keep us posted about the progress. 

 

Concerning why you shouldn't use a HW Raid in combination with Truenas, the official ZFS documentation has a great section about this you find it here: https://openzfs.github.io/openzfs-docs/Performance and Tuning/Hardware.html#hardware-raid-controllers

[Sam of Ham]

This is with the B120i disabled and all 4 drives removed (just the boot drive) in Bay 5:

This is with the B120i enabled and all 4 drives removed (just the boot drive) in Bay 5:

After cleaning the disk (just by pulling it out, dropping it into Windows and doing dispart clean to wipe the parts) 

I don't think I can disable the B120i because I think it needs to be there to mark Disk 5 as bootable... Unless I'm crazy. (I'm probably crazy.) 

Whatever I do, I can't get anything to boot anymore! Losing my shit, haha! 

Thanks for your help so far @PsychoWards!!

[PsychoWards]

The above messages are coming after selecting the boot drive where Hexos should be installed or beforehand?