Lenovo RD650 Reboot Loop at loading initial ramdisk - TrueNAS-22.12.4.2 -> TrueNAS-23.10.2

[IkeUrb]

Tonight I attempted to upgrade from TrueNAS Scale 22.12.4.2 -> TrueNAS-23.10.2

Upon boot, the system continuosly reboots at Loading Initial Ramdisk.

At grub, I tried booting is safe mode, but same results.

So, I was forced to load 22.12.4.2 kernel from Grub menu. TrueNAS Scale boots and works great from this kernel.

Not sure where to go from here. Any help is appreciated.

Thanks in advance.

 

[artlessknave]

hardware? just a model number is not your actual hardware, the best that tells us is a broad range of hardware you might have, and requiring us to go learn about a whole server to even attempt a guess.

 

[IkeUrb]

Sorry, it was late in the night and I was anxious to get some help. This is my home machine as a hobby and learning platform. My first post here on this forum.

The machine is a 1U server with the following:

OS Version: TrueNAS-SCALE-22.12.4.2

Server: Lenovo ThinkServer RD650
Motherboard PN:SB20A05917

CPU: Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz

Memory:
Manufacturer Name: Micron Technology
Memory PN: 18ASF1G72PZ-2G1A2
Memory Frequency: 1866MHz
16 GB

Drives:
Onboard RAID Controller
Boot Drive is SSD HPS750 256GB
Four 4 TB Seagate IronWolf ST4000VN006 hard drives mirrored VDEV storage

Network Card: Onboard Lenovo NIC

 

[artlessknave]

hmm. i dont see anything glaringly wrong....

Onboard RAID Controller

I do question this, however. im not sure it would be causing the boot loops, but it's concerning to see "raid controller".
it appears to be the same raid controller series in this thread.

SMART functionality on ThinkServer RAID 720i AnyRAID (smartd.conf is empty)

Please take into account before reading I am a complete FreeNAS noob. So I have a thinkserver TD350 that has a ThinkServer RAID 720i AnyRAID (LSI 3108) raid controller installed . Let me get the other hardware in here too. The CPU is a single Xeon E5-2640 v4, It has 48gb off ddr3. I don't know...

www.truenas.com

one thing many of us do is put system info into the forum signature, then its available with every post.

 

[IkeUrb]

I am not using the native onboard RAID configuration. I have the drives installed in the server and configured the mirror setup with TrueNAS.

Not sure, but I think that mitigates any issue with compatibility/incompatibility with RAID controllers.

Thanks for the tip regarding system information set as part of the forum signature. I'll work on that now.

 

[IkeUrb]

Also, TrueNAS was booting and working fine on 22.12.4.2.
The boot loop started when I attempted to update to 23.10.2.

Manually booting from Grub with 22.12.4.2 allows the machine to boot fine as prior to the upgrade attempt. So, I removed the instance of 23.10.2 through the TrueNAS UI and set 22.12.4.2 to the default version until we find a solution.

 

[artlessknave]

Not sure, but I think that mitigates any issue with compatibility/incompatibility with RAID controllers.

no, it doesn't you need to have the drives running in an IT mode. sometimes called passthrough. some raid controllers have "JBOD" mode but those are are not all the same. some people also try to make every disk a 'raid0' but this is also a terrible idea.

again, I'm not sure that would be the cause of the boot loops, but its also possible that, if you are using the raid driver, something changed in the boot image and it no longer works. raid cards also tend to have cache and that can cause issues with zfs as well. there are a plethora of reasons to make sure you do have have any RAID abstraction between zfs and the disks, many of them are outlines in that post and the other mountains of posts jgreco has made.
one of the common issues is that these pseudo raid setups 'work' until one day they just don't.
if any of that is the case then the fix is to get a proper HBA in and eliminate the raid card being an issue.