Re: General Protection Fault in md raid10

[Colgate Minuette <rabbit@minuette.net>]

On Sunday, April 28, 2024 3:25:29 PM PDT Roman Mamedov wrote:
> On Sun, 28 Apr 2024 15:16:27 -0700
> 
> Colgate Minuette <rabbit@minuette.net> wrote:
> > I just tried RAID10 on the same HBA/cables with 4 seagate 4TB SAS HDDs,
> > and it is functioning correctly. Syncing correctly and able to write/read
> > from the md device.
> 
> With those 15 TB SSDs, maybe something wonky with the large size?
> 
> Did you try creating a smaller partition on each, maybe just start with 4,
> to not redo all of them, since you say 4 also repros the issue. Test with a
> 4 TB or even 1TB partitions as RAID members.

Created a 2TB partition on 4 of the drives, created md RAID10 on the 2TB 
partitions, and got another protection fault shortly after starting the array.

[  515.504412] md/raid10:md51: not clean -- starting background reconstruction
[  515.504414] md/raid10:md51: active with 4 out of 4 devices
[  515.530362] md51: detected capacity change from 0 to 8388079616
[  515.530445] md: resync of RAID array md51
[  524.083652] general protection fault, probably for non-canonical address 
0xb1c8a7fff899a: 0000 [#1] PREEMPT SMP NOPTI


-Colgate