Something else that may be worth looking at - Don't know what connector your using from external SAS port to the JBOD, but maybe it needs checking/reseating/replacing, I'm thinking some sort of short connection there is causing all of this PCI monkey-business.
[SOLVED] HBA Card/JBOD issues
- Thread starter Tezza
- Start date
-
- Tags
- hba jbod pve-no-subscription
I have reseated the cable. But not replaced it. i have a cable that i could try replacing it with when i get home today.
Aside from that cable, when you get home today check if it's definitely booting into the older 6.5 kernel instead of the new default of 6.8.
When it is, I'd try grabbing the kernel logs again, but for the 6.5 kernel this time.
That problem showing up in the kernel is pretty much the lpfc driver crashing as it's going about it's initialisation routine, and is a serious issue.
Hopefully a simple cable change fixes it.
When it is, I'd try grabbing the kernel logs again, but for the 6.5 kernel this time.
That problem showing up in the kernel is pretty much the lpfc driver crashing as it's going about it's initialisation routine, and is a serious issue.
Hopefully a simple cable change fixes it.
As a diagnostic step, it might be worth running memtest86+ on the system for a full pass as well, just to eliminate one possibility from the list of potential causes:
https://github.com/memtest86plus/memtest86plus/releases
https://github.com/memtest86plus/memtest86plus/releases
So i just got home and i was curious what happened and it never rebooted. I turned on the monitor to that server and it still said "rebooting..."
so i thought there could be hope. I manually turned the server off then back on and it used the 6.2 kernel i know has worked before and it works just fine
i have network and i can see my jbod drives.
so i thought there could be hope. I manually turned the server off then back on and it used the 6.2 kernel i know has worked before and it works just fine
i have network and i can see my jbod drives.
Last edited:
That's weird.
Would you be ok to paste the output from
uname -a here?As an example, here's the output from one of my systems:
Bash:
# uname -a
Linux home1 6.5.13-5-pve #1 SMP PREEMPT_DYNAMIC PMX 6.5.13-5 (2024-04-05T11:03Z) x86_64 GNU/LinuxThat being said, since things are working with the 6.2 kernel it's almost definitely nothing hardware (cable, etc) related.
Linux clotho 6.2.16-20-pve #1 SMP PREEMPT_DYNAMIC PMX 6.2.16-20 (2023-12-01T13:17Z) x86_64 GNU/Linux
I can add 6.5 and see if it works i just had 6.2 installed and wanted to try it since i knew it worked with my JBOD before.
Yeah, I'd try 6.5 as an experiment to see whether or not it works.
While technically you can stick with 6.2 "forever", at some point in the future it'll become so old that problems start to arise. That's possibly years away though.
If we can figure out a working solution with the more standard 6.5 kernel (and maybe 6.8 after they fix the many reported bugs), that would probably be a better future-proofing path. That's in theory anyway.
While technically you can stick with 6.2 "forever", at some point in the future it'll become so old that problems start to arise. That's possibly years away though.
If we can figure out a working solution with the more standard 6.5 kernel (and maybe 6.8 after they fix the many reported bugs), that would probably be a better future-proofing path. That's in theory anyway.
6.5 works thank you all. First time using the forums. The proxmox community is amazing thank you all for the help.
As I imagined (my post above) this is yet another one of those kernel 6.8 bugs! I hope all these bugs get ironed out / mitigated in a future release.
I'm happy you've got yourself worked out - although you'll obviously be waiting for the next kernel update (which you should try when it becomes available - you can always go back again to the working 6.5 one again).
Maybe tag mark the thread-title with [SOLVED], (upper right hand corner under title).
I'm happy you've got yourself worked out - although you'll obviously be waiting for the next kernel update (which you should try when it becomes available - you can always go back again to the working 6.5 one again).
Maybe tag mark the thread-title with [SOLVED], (upper right hand corner under title).