[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: hibernate and swap partition size (newbie question)

On Fri, Feb 22, 2008 at 7:52 AM, Kamaraju S Kusumanchi
<kamaraju@bluebottle.com> wrote:
> Jimmy Wu wrote:
>  >>From what I've read online, I get the general idea that in order to be
>  > able to hibernate/suspend to disk properly, the swap partition has to
>  > be big enough to hold all of the RAM inside it, right?
>  >
>  > Is it possible to hibernate if my swap partition is smaller than my
>  > RAM?  I have 2 GB of RAM, and when I installed Debian, I figured I
>  > would hardly ever need that much, so I made swap 1.4 GB.
>  >
>  Are you aware that you can resize your partitions non destructively using
>  something like qtparted? First backup all your data before you do anything
>  like this. This is what I did when I found out that my RAM size is larger
>  than my swap partition.

I always thought resizing or doing any partition editing carried some
risk of losing data (ie no guarantees), but perhaps ext3 is different.

Anyways, I think I have ruled out the low swap explanation:

On Fri, Feb 22, 2008 at 2:19 AM, Selim T. Erdogan <selim@cs.utexas.edu> wrote:
> On Fri, Feb 22, 2008 at 12:11:54AM -0500, Jimmy Wu wrote:
>  >
>  > Not really what you were saying, but I suppose it might work.  But
>  > first I have to figure out if it really is inadequate swap that's
>  > giving me grief.
>  I would assume that upon doing a fresh boot-up you would be using much
>  less memory than your available swap partition.  (You can check and
>  confirm using, say, top.)  Then if you try suspending and still have the
>  same problems when resuming, I think it would be a good indicator that
>  your problems lie elsewhere.

Right after boot, I logged in to tty1 and did a sudo pm-hibernate
While staring at the screen for messages, I noticed that the snapshot
image was less than 400 MB, and it correctly determined free swap as
just short of 1.4 GB - so swap is more than enough.

On resume, I got a lot of beeps, but after waiting for like 2-3
minutes or so, I was back at my console prompt.  However, when I tried
to switch over to gdm on tty7, the screen is black, system goes
unresponsive, and I can't get back to my tty1 anymore.

Just as an experiment, I did a sudo hibernate -v3 > hibernate.out, and
it says that it was unable to unload nvidia and aborts hibernation
(see attached file).  So I guess pm-hibernate kind of went ahead and
shut down without properly taking care of nvidia, so that is why I had
an unresponsive X server on resume, right? If that's what's happening,
is there any way to get nvidia properly unloaded?  I am running a
stock kernel, and have nvidia installed from the Debian repositories.

Also, there's a script in this article I found:
http://www.linux.com/feature/114220.  I never like running scripts
that I don't understand, and I was wondering if whatever it's doing
with the video card solve my problem?

Thanks again for everyone's help and responses
Jimmy Wu
Registered Linux User #454138

Attachment: hibernate.out
Description: Binary data

Reply to: