Page 2 of 2
Re: Is the SATA controller *supposed* to work?
Posted: 21. Jan 2011, 22:02
by DanielSmedegaardBuus
Perryg wrote:Boots from CD > I choose "Boot from first harddisk" > GRUB shows > hanging... > Get "Gave up waiting for root device", as shown before
And what happens if you remove the CD at this point instead of booting from the CD and let it boot direct from the HDD?
That's the "I shut down machine > Settings > I remove the CD image > OK > Start > same thing..." part
Did you try the rapidshare 7z file? Does it behave the same on your computer?
Re: Is the SATA controller *supposed* to work?
Posted: 21. Jan 2011, 22:35
by Perryg
It's still downloading. really slow
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 00:33
by Perryg
Your install is corrupt. Grub is not installed properly and other missing modules.
So I installed a new copy of Ubuntu 10.04 Server 32-bit using SATA so you can see that it does in fact work.
- Userv-10.04-SATA.png (20.52 KiB) Viewed 3892 times
I think I can repair your copy but really don't see the point, since you would need to change things all over again anyway.
So what is Drive E? is it another drive or an external drive, if external is it eSata, or USB?
Is the host up-to-date and which SP are you using? Also if it a full install, no nLite, or hacked versions.
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:14
by Perryg
You know it looks like you didn't install the bootloader. Are you sure that when you were asked (toward the end) you said yes?
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:15
by DanielSmedegaardBuus
Perryg wrote:Your install is corrupt. Grub is not installed properly and other missing modules.
So I installed a new copy of Ubuntu 10.04 Server 32-bit using SATA so you can see that it does in fact work.
Userv-10.04-SATA.png
I think I can repair your copy but really don't see the point, since you would need to change things all over again anyway.
I think you just proved my previous theory: that it's just modules missing. I installed using the "Install a minimal virtual machine" option, which is
supposed to optimize for being virtualized, which gives the virtual kernel while yours is the generic pae kernel. Most likely, the pae one or one even more generic is also the one used during install which is why the SATA controller is accessible during install but not after. I'm gonna try apt-getting the generic kernel, and I'll probably end up with functional SATA.
Will report back
Perryg wrote:So what is Drive E? is it another drive or an external drive, if external is it eSata, or USB?
Is the host up-to-date and which SP are you using? Also if it a full install, no nLite, or hacked versions.
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:16
by DanielSmedegaardBuus
Perryg wrote:You know it looks like you didn't install the bootloader. Are you sure that when you were asked (toward the end) you said yes?
Yup, it's GRUB giving errors in the first place
Booting the VM now - takes a reaaaally long time after I added the 12 raw drives!
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:17
by Perryg
virtual kernel probably will not work in a VirtualBox guest. VirtualBox does not support a virtualizer inside a virtualizer.
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:25
by Perryg
Well your subject line indicates that you felt this was a VirtualBox issue, along with some of your posted responses, which it isn't. Now since I have spent the time to actually show you what it was and what needed to be done was actually outside the scope of VirtualBox maybe the next time you have an issue you can word it a little better instead of jumping to judgment.
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 01:32
by DanielSmedegaardBuus
Perryg wrote:virtual kernel probably will not work in a VirtualBox guest. VirtualBox does not support a virtualizer inside a virtualizer.
It's not a virtualizer. It's a kernel aimed at virtualized systems, i.e. guest systems.
Just rebooted with the pae kernel on SATA - works. So, it is a module missing.
-virtual modules:
Module Size Used by
vboxsf 35783 1
ppdev 5556 0
parport_pc 26378 0
psmouse 59033 0
serio_raw 4022 0
vboxguest 164940 2 vboxsf
lp 7342 0
parport 31492 3 ppdev,parport_pc,lp
mptspi 14888 2
mptscsih 32294 1 mptspi
mptbase 86626 2 mptspi,mptscsih
e1000 97749 0
-generic-pae modules:
Module Size Used by
ppdev 5556 0
snd_intel8x0 25632 0
parport_pc 26378 0
snd_ac97_codec 99227 1 snd_intel8x0
ac97_bus 1014 1 snd_ac97_codec
snd_pcm 71603 2 snd_intel8x0,snd_ac97_codec
snd_timer 19067 1 snd_pcm
psmouse 59033 0
joydev 8767 0
snd 49038 4 snd_intel8x0,snd_ac97_codec,snd_pcm,snd_timer
serio_raw 4022 0
soundcore 880 1 snd
snd_page_alloc 7216 2 snd_intel8x0,snd_pcm
i2c_piix4 8795 0
lp 7342 0
parport 31492 3 ppdev,parport_pc,lp
usbhid 36978 0
hid 67742 1 usbhid
mptspi 14888 0
e1000 97749 0
mptscsih 32294 1 mptspi
ahci 19198 2
mptbase 86626 2 mptspi,mptscsih
libahci 21728 1 ahci
The ones in bold seem interesting. Gonna hit the sack now, and I'll try adding them to the -virtual kernel tomorrow, see what happens. Gonna open a ticket with ubuntu, too. Clearly this is not right.
Goodnight
Re: Is the SATA controller *supposed* to work?
Posted: 22. Jan 2011, 15:11
by DanielSmedegaardBuus
Funny, trying to insmod and modprobe my way to happiness with the -virtual kernel led me to error messages which led me to the launchpad bug report for this exact issue:
https://bugs.launchpad.net/ubuntu/maver ... bug/570542
Seems the bug was introduced with Lucid (correlates with my identical experiences last year with 10.04), fixed for Maverick alpha-2, then re-appeared in the final ISOs, and persists up to the latest kernel version in the repos.
Anyone wanting to contribute getting this bug fixed ASAP should go to that URL, sign yourself up, and add yourself as an affected person.
Cool that it's an Ubuntu bug, not a VBox one. This new V4 really rocks - and it seems much more stable than the 3.x branch.
Cheers for the sparring!
EDIT: Wrong URL
Re: Is the SATA controller *supposed* to work? [RESOLVED]
Posted: 22. Jan 2011, 16:10
by Perryg
IIRC the virtualized kernel is geared towards the type 1 hypervisors (xen and the likes) *not* type 2.
Re: Is the SATA controller *supposed* to work? [RESOLVED]
Posted: 22. Jan 2011, 17:08
by DanielSmedegaardBuus
Perryg wrote:IIRC the virtualized kernel is geared towards the type 1 hypervisors (xen and the likes) *not* type 2.
Oh no, the -virtual kernel is not for host systems, i'ts aimed at
virtualized guests. The -virtual kernel is - among other things - built with PAE (
http://en.wikipedia.org/wiki/Physical_Address_Extension), which makes it possible to enable PAE/NX in VBox (and VMware). When a system is given a kernel that is aware of being virtualized, you can seriously optimize things in ways you couldn't without. And, you can omit a boatload of other kernel modules since you have a fairly good idea of what hardware you DON'T have to support
I.e. a lean, mean virtualized machine.
If you're serious about performance (and don't mind manually choosing what extra packages to add when installing *buntu), it's highly recommended to use the "Install a minimal virtual machine" option when installing in VBox or VMware. You'll notice the difference right away when booting. It's blazing
Re: Is the SATA controller *supposed* to work? [RESOLVED]
Posted: 22. Jan 2011, 17:12
by DanielSmedegaardBuus
Actually, I don't know if the "if you don't mind manually selecting packages" caveat I mention applies to desktop versions of *buntu, I've only installed server versions myself. Worth a test!