Supermicro X8STI woes
- Thread starter 3rased
- Start date
yes i have a 3ware raid card that picks up fine but i want to use the onboard sata to attach a 3rd hard drive as a backup volume
the bios of the machine sees the drive but proxmox doesn't see it i wonder if this is a kernel thing where the right drivers aren't loaded
the bios of the machine sees the drive but proxmox doesn't see it i wonder if this is a kernel thing where the right drivers aren't loaded
take a look in the bios and see if there is a selection how to use the sata, there should be something like:
- raid
- ahci
- ide
i dont have physical access to the machine (as its a dedicated server) but am getting the tech support guys to check this out now
we rebooted to try this but on reboot we got a kernel panic
our raid unit is a 3ware raid with 2x 1.5TB sata drives connected in mirror arrangement we were trying to get an onboard sata interface to pick up an extra 500gb hd to be mounted for backup (the onboard hd is plugged into socket 4 on the mobo - would be secondary slave in ide mode)
Code:
switch_root: can't open '/dev/console' : No such file or directory
Kernel panic - not syncing: attempted to kill init!our raid unit is a 3ware raid with 2x 1.5TB sata drives connected in mirror arrangement we were trying to get an onboard sata interface to pick up an extra 500gb hd to be mounted for backup (the onboard hd is plugged into socket 4 on the mobo - would be secondary slave in ide mode)
Last edited:
to fully respond to my previous message
it was actually in enhanced IDE mode in the bios and proxmox didnt pick it up
i switched to AHCI, proxmox did pick it up but i got kernel panic as above - i did notice that my raid drives had moved from sda to sdb
i switched to raid mode and i got the same result
compatible AHCI mode - kernel panic was picked up
compatible IDE mode - kernel panic not picked up as far as i could see
compatible RAID mode - kernel panic was picked up
only working mode for me was enhance IDE at which point proxmox did not pick up the onboard drive
it was actually in enhanced IDE mode in the bios and proxmox didnt pick it up
i switched to AHCI, proxmox did pick it up but i got kernel panic as above - i did notice that my raid drives had moved from sda to sdb
i switched to raid mode and i got the same result
compatible AHCI mode - kernel panic was picked up
compatible IDE mode - kernel panic not picked up as far as i could see
compatible RAID mode - kernel panic was picked up
only working mode for me was enhance IDE at which point proxmox did not pick up the onboard drive
your gonna get the same problems i did
now if only we could make the onboard drive stay last as the kernel panics i believe are caused by the fact it shifts your RAID unit to sdb from sda, i dont know if there is anyway of resolving this without a reinstall with both drives attached
for reference you need to have the bios in Enhanced AHCI mode for proxmox to pick up the drive but then you get kernel panic from proxmox
i suspect this has something to do with grub or is grub related
now if only we could make the onboard drive stay last as the kernel panics i believe are caused by the fact it shifts your RAID unit to sdb from sda, i dont know if there is anyway of resolving this without a reinstall with both drives attached
for reference you need to have the bios in Enhanced AHCI mode for proxmox to pick up the drive but then you get kernel panic from proxmox
i suspect this has something to do with grub or is grub related
Last edited:
only solution i know of currently is to reinstall with the hard drive active (enhanced AHCI) and make sure you get the right volume to install too (most likely sdb)
further update
its something to do with the references of the hd changing from sda to sdb and trying to change into the root of something that is not there
the part of the boot script that is causing the issue: change root
its trying to change the root to /dev/mapper/pve-root (which is normal) now for some reason its reference (i assume symlink) is the old one (sda) which is causing the issue
its something to do with the references of the hd changing from sda to sdb and trying to change into the root of something that is not there
the part of the boot script that is causing the issue: change root
its trying to change the root to /dev/mapper/pve-root (which is normal) now for some reason its reference (i assume symlink) is the old one (sda) which is causing the issue
further update
its something to do with the references of the hd changing from sda to sdb and trying to change into the root of something that is not there
the part of the boot script that is causing the issue: change root
its trying to change the root to /dev/mapper/pve-root (which is normal) now for some reason its reference (i assume symlink) is the old one (sda) which is causing the issue
its something to do with the references of the hd changing from sda to sdb and trying to change into the root of something that is not there
the part of the boot script that is causing the issue: change root
its trying to change the root to /dev/mapper/pve-root (which is normal) now for some reason its reference (i assume symlink) is the old one (sda) which is causing the issue
/dev/mapper/pve-root is an LVM volume, which is detected by scanning the contend of the disk.
yeah i had worked that much out, really strange why adding the extra HD causes the mapping to go wrong and why it doesnt rescan (its like the scan gets interupted)
anyway i solved my problem - i had to reinstall, annoyingly because there is no way of doing a PXE install of proxmox i had to remotely PXE lenny 64bit and manually install PVE on top, but it went fine without a hitch with the exception that debian has decided to rename my network interfaces from eth0 and eth1 to eth2 and eth3 respectively, it seems persistant but still not sure about it.
just got fsck issues to contend with at the moment, but seems fine
anyway i solved my problem - i had to reinstall, annoyingly because there is no way of doing a PXE install of proxmox i had to remotely PXE lenny 64bit and manually install PVE on top, but it went fine without a hitch with the exception that debian has decided to rename my network interfaces from eth0 and eth1 to eth2 and eth3 respectively, it seems persistant but still not sure about it.
just got fsck issues to contend with at the moment, but seems fine