Installing TuxOnIce on CentOS Rocks cluster distribution

_______________________________________________
TuxOnIce-users mailing list
TuxOnIce-users <at> lists.tuxonice.net
http://lists.tuxonice.net/listinfo/tuxonice-users

Re: Installing TuxOnIce on CentOS Rocks cluster distribution

On 3 Feb 2012, Mämmelä Olli verbalised:

> I am planning on installing TuxOnIce to CentOS Rocks Linux 5.4.3
> (Viper). The Kernel information is 2.6.18-238.19.1-el5 x86_64. Is it

Nobody maintains ToI for such kernels (being as it is based on an
ancient baseline kernel and then has a large number of patches for later
kernels applied, some very invasive, with all the patches glumped
together in a single huge patch specifically to make it hard to figure
out what's been done).

Even if a patch existed against 2.6.18 (which was released almost six
years ago), and even if it applied against the heavily modified RHEL
version, it would be guaranteed to have large numbers of bugs fixed in
ToI releases for later kernels, since there's no way it would be tested
very often. Worse yet, TuxOnIce depends on the freeze/hibernate
machinery in the kernel proper, which was in distinctly embryonic state
that long ago: and RHEL is really meant for always-on servers, so the
likelihood of RH having backported later suspend/resume bugfixes into
their RHEL kernels is distinctly remote.

My advice is that if you're planning to use a desktop, you should
probably use a more desktoppy OS than CentOS. (That would be pretty much
*any* Linux distro not based on RHEL.)

--

-- 
NULL && (void)
_______________________________________________
TuxOnIce-users mailing list
TuxOnIce-users <at> lists.tuxonice.net
http://lists.tuxonice.net/listinfo/tuxonice-users

Re: Installing TuxOnIce on CentOS Rocks cluster distribution

On Fri, Feb 03, 2012 at 07:52:21PM +0000, Nix wrote:
> On 3 Feb 2012, Mämmelä Olli verbalised:
> 
> > I am planning on installing TuxOnIce to CentOS Rocks Linux 5.4.3
> > (Viper). The Kernel information is 2.6.18-238.19.1-el5 x86_64. Is it
> 
> Nobody maintains ToI for such kernels (being as it is based on an
> ancient baseline kernel and then has a large number of patches for
> later kernels applied, some very invasive, with all the patches
> glumped together in a single huge patch specifically to make it hard
> to figure out what's been done).

Older Tuxonice-builds for CentOS/RHEL5 are available at
http://mhensler.de/swsusp/download_en.php (or http://atrpms.net/). The
last released kernel is 2.6.18-92.1.6.el5 (with TuxOnIce 2.2.9).

I will try to build the current kernel with the old patchset. If that
works without to much hassle you can expect to find updated RPMs on the
above pages within 1 week or so.

Regards,
Matthias

_______________________________________________
TuxOnIce-users mailing list
TuxOnIce-users <at> lists.tuxonice.net
http://lists.tuxonice.net/listinfo/tuxonice-users

Re: Installing TuxOnIce on CentOS Rocks cluster distribution

On Sat, Feb 04, 2012 at 02:57:23PM +0100, Matthias Hensler wrote:
> I will try to build the current kernel with the old patchset. If that
> works without to much hassle you can expect to find updated RPMs on
> the above pages within 1 week or so.

The current Redhat Enterprise 5 / CentOS kernel (2.6.18-274.17.1) with
TuxOnIce-support is now available from http://mhensler.de/swsusp/ and
should be available on ATrpms soon.

Greetings,
Matthias

_______________________________________________
TuxOnIce-users mailing list
TuxOnIce-users <at> lists.tuxonice.net
http://lists.tuxonice.net/listinfo/tuxonice-users

...it cannot sleep...

Hi,

I am using GENTOO, which I update about once a day.

I have installed sys-power/hibernate-script-2.0-r1

As source for this package 
Homepage:            http://www.tuxonice.net/
is mentioned.

My system consists of:

 Kernel > Linux 3.2.5 (vanilla kernel) 
 CPU    > AMD Phenom(tm) II X6 1090T Processor AuthenticAMD GNU/Linux
 MOBO   > ASUS Crosshair Formula 4 
 GRAKA1 > VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 430] (rev a1)
 GRAKA2 > VGA compatible controller: NVIDIA Corporation GF110 [GeForce GTX 560 Ti] (rev a1)

The GT430 Graka is used for desktop perposes - it is connected to my
monitor. The GTX 560 Ti is used for rendering purposes only. It is not 
connected to any monitor.

As I had installed only one of the Graka, there were no problems to
hibernate the desktop pc. This was valid for both cards.

As I had installed both cards to not to "lock" my desktop while
rendering, peace was over...no sleep was the result.

When I try to hibernate my PC to disk, the script runs until the
monitor goes black. But the computer does not shut off and when I
reboot it, it does not behave as it had been hibernted.

Since the script runs until the monitor blanks I see no chance to get
hands on the last lines of the output of the script.

What can I do to re-enable the hibernation capabilities of my PC?

Thank you very much of any help in advance!

Best regards,
mcc

Patch for the kernel 3.2 available?

Hello,

is there a "ready to use" patch available for the latest 3.2 kernel
sources (like 3.2.5)?

Regards

Newbie questions

Hi,

Thanks to Nigel and others for all their work on TuxOnIce.

I've been using in-kernel suspend for some time on my Debian stable box, 
but have become frustrated with the long time taken to resume, so have 
been trying TOI. Kernel is 2.6.32 and latest stable TOI (3.2). I am 
perhaps using a more unusual configuration, / is separate non-LVM 
partition, formatted in xfs, system and home folders are xfs on LVM, 
swap is separate non-LVM partition. I use an initramfs and grub2.

I followed the instructions from 
http://r.outlyer.net/debian:toikerneldeb (which mostly worked) as being 
the best I could find for my system. I also read the Arch and Slax Wikis.

It'd be great if the ToI Wiki could be fixed, so there were some updated 
fuller instructions available.

Do you recommend using the hibernate script?

I wondered if someone could post a typical config file(s) for with and 
without the hibernate script?

I have had fast and successful suspends and resumes, but also some 
issues...

Usually suspends, but sometimes won't resume.

Is there any more info on how to debug?

I tried to install an Ubuntu ToI kernel from the PPA for Lucid (quite 
similar to Debian Squeeze), but it failed due to missing dependencies.  
I also looked at Liquorix, but they are too new for my system. Any 
thoughts? It'd be great if the TOI-patch package that is in Debian could 
be updated and got into stable...

What memory size does 'pages' refer to in the context of the swap space? 
The value given there is less than I'd expected, and is less than the 
'memory needed' estimate for the image, but that is in bytes. I have 4G 
RAM, so am decided to run the pae kernel, marginal though I suspect. I 
have ~15GB swap.

Can the 'expected compression percentage' be set permanently somewhere, 
and should it need to be? Mine is zero?

LZF should be used rather than LZO for 2.6.32?

Thanks for now,
A