PCLinuxOS Openbox Bonsai 2010.11<s>.13</s> who wants to test ?

[loukingjr]

btw? I just noticed, my sound is always starting up "muted". is there a fix?

thanks

Yes, install alsa-utils and configure your sound.

PS: if you install volumeicon, uncomment it's line in .config/openbox/autostart.sh, reload the session, then you will have volumeicon instead of the lxpanel plugin. Then, if you install gnome-alsamixer, you can change the command which is in the preferences of Volumeicon : the default is «xterm -e 'alsamixer'», you can replace it with just «gnome-alsamixer», it will start... gnome-alsamixer instead of alsamixer in a console.

thanks hun. enjoy your holiday.

[melodie]

thanks hun.

How did you know my ancesters were the huns ?

enjoy your holiday.

Thanks, same to you !

[dalrun]

I don't see a more recent topic. So far this has been a good replacement for KDE Minime 2009 ('10 and all KDE4 live CD's stall on this PC). Thank you.

What's bugging me is the slow bootup due to 'could not (resolve) resume device' sda5/swap. I don't really understand ramzswap, is a swap partition not needed/used? Everything seems to be working fine. What is the best way to handle this - ramz_backing_swap="/dev/sda5"? I also get an apparently harmless 'libpng error: read error'.

[melodie]

I don't see a more recent topic. So far this has been a good replacement for KDE Minime 2009 ('10 and all KDE4 live CD's stall on this PC). Thank you.

What's bugging me is the slow bootup due to 'could not (resolve) resume device' sda5/swap. I don't really understand ramzswap, is a swap partition not needed/used? Everything seems to be working fine. What is the best way to handle this - ramz_backing_swap="/dev/sda5"? I also get an apparently harmless 'libpng error: read error'.

Hi,

'libpng error: read error' occurs in all versions. Neal tried to solve it, Texstar tried an update of png, nothing helped. A message such as "'could not (resolve) resume device' sda5/swap" is not related to ramzswap, but to the way the initrd is constructed, and occurs if you have moved partitions, the place of the partition is "marked" with it's uuid... something like that. The easiest way to solve it is to install another kernel, and reboot to it. A new initrd will be created and you will get rid of this message and of the boot delay at same time.

What is ramzswap ? Explanation at compcache website:
http://code.google.com/p/compcache

It will next be replaced by a simple configuration that will make use of zram, a kernel module. For the one kernel we currently have, to be able to make use of Compcache and it's module ramzswap we need the additional program rzscontrol, in order to use it from boot to the end of the session, in a way that can fit all machines, (PC architecture of course).

The goal is to get a memory potential bigger than what the available ram really provides.

It is done in a way such as using it preferably before the swap partition be used.

some interesting discussions here:
http://code.google.com/p/compcache/wiki/zramperf

ie:

Even if zram's overhead is between 50% and 70%, it is still much faster than disk access.

[dalrun]

I didn't change anything after the install, so I must have goofed with the custom partitioning (small disk, / and swap). Changing kernels and removing the 'append uuid...' command from grub did the trick. Two things puzzle me, the swap partition doesn't get mounted (mounting it manually works) and I also got a 'not a block device, skipping' message when installing the kernel. Are either of those related to compcache?

...I'm not positive, but it looks like the fstab swap UUID is the same one I've been having grief with. Changing kernels just removed the swap mount call?

[melodie]

'not a block device, skipping' message when installing the kernel.

Hi,

The message "as is" ? nothing more ?

My swap partition is mounted automatically from fstab. It's on /dev/sda7 and the line in fstab is so:

Code: [Select]

/dev/sda7 swap swap defaults 0 0
do you have something similar ?

If you suspect compcache of anything, deactivate it and try without it.
You need only to do 2 things for that : go as root in terminal in /usr/local/sbin, and rename it or remove execution from rzscontrol, and in /etc/rc.d/rc.local, comment the related lines:

######### COMPCACHE ##########

# Name of the ramzswap device (/dev/ramzwap0 is the default).
RAMZ_DEVICE="/dev/ramzswap0"

# Ramzswap will act like a swap front end if RAMZ_BACKING_SWAP is defined.
# Writes are forwarded to this device when memory limit is reached or data
# is not compressible.
# i.e. RAMZ_BACKING_SWAP="/dev/sda3"
RAMZ_BACKING_SWAP=""

# If you have several swap devices (or partitions), you may prefer to have
# a higher priority for ramzswap.
RAMZ_SWAP_PRIORITY=30

# You may want to change this if you want to add some extra parameters to
# rzcontrol (see http://code.google.com/p/compcache/wiki/CompilingAndUsingNew)
RAMZ_CMD="--init"


######## NOTHING TO CHANGE BELOW THIS ##########

modprobe ramzswap

if grep -wq "$RAMZ_DEVICE" /proc/swaps; then
   # I met with some kernel issues while resetting ramzswap devices.
   # Normally, this shouldn't be executed if you don't play with rc.local.
   swapoff "$RAMZ_DEVICE"
   /usr/local/sbin/rzscontrol $RAMZ_DEVICE --reset
fi


if [ -n "$RAMZ_BACKING_SWAP" ] ; then
   if grep -wq "$RAMZ_BACKING_SWAP" /proc/swaps; then
      swapoff "$RAMZ_BACKING_SWAP"
   fi
   RAMZ_CMD="--backing_swap $RAMZ_BACKING_SWAP $RAMZ_CMD"
fi

/usr/local/sbin/rzscontrol $RAMZ_DEVICE $RAMZ_CMD
swapon $RAMZ_DEVICE -p $RAMZ_SWAP_PRIORITY

######### /COMPCACHE ##########[

Or rename it to "rc.local-back", and make a new one, containing only the part which is before this one.

But I am pretty sure there is no wrong effect coming from it. Maybe the swap partition of your's is rotten and reformatting it would correct it.

Changing kernels just removed the swap mount call?

No, it did a new initrd. The initrd's in PCLinuxOS contain an information about what and where the partitions are. This is annoying by the way, when we change a hard drive from one machine to another, or when we move partitions around, and reformat some of them. The initrd has to be redone. To redo the initrd, some command lines exist. Installing a new kernel is the fast/lasy/easy method.

[Was_Just19]

mkinitrd     seems to be command ......  see the 'man' page for the options or just use the defaults as is implied by the example which can be seen by typing
mkinitrd --help      in a terminal (as root).

man:mkinitrd     in Konqueror to get the full man page in a very legible format.

[dalrun]

'not a block device, skipping' message when installing the kernel.

Hi,

The message "as is" ? nothing more ?

Yes, that's basically all Synaptic said, no details.

My swap partition is mounted automatically from fstab. It's on /dev/sda7 and the line in fstab is so:

Code: [Select]

/dev/sda7 swap swap defaults 0 0
do you have something similar ?

My fstab has UUID's

Code: [Select]

# Entry for /dev/sda1 :
UUID=b49420d6-d48b-11db-acea-939d80333f57 / ext3 defaults 1 1
none /proc proc defaults 0 0
# Entry for /dev/sda5 :
UUID=cb6f6cfc-d48b-11db-ab3b-255d36b591f8 swap swap defaults 0 0
none /dev/pts devpts defaults 0 0

I've been running the system without sda5/swap mounted (it also runs fine w/ swap mounted) and, without exploring the details, compcache seems to be working fine. It's currently using 3 out of 93MB. Since I've never had a partition issue it was hard not to associate my swap partition issues with this newfangled compcache.

My point about the new initrd was that the only apparent change to the system was that bootup no longer tried to do anything with sda5/swap - whatever that message '...<partition #> (one minute timeout)' is that resulted in the 'could not (resolve) resume device UUID' message. The question mark was because I didn't know if 'mount call' was an appropriate name for it (lousy grammar, I know).

[melodie]

Hi,
My English is not perfect either, my native language being French. Is your problem solved now ? If you want to check the UUID numbers, you can type as root "blkid", which will list all partitions along with their UUID.