problem with persistence

Post Reply
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

problem with persistence

Post by mimosa »

Now I'm inside it, it's great! I particularly like the idea of the personalisable USB, where I can put my wireless configuration, Adblock plus, etc. etc. :)

However, I am having some problems with persistence. The machine I'm testing the new Salix live out on has Ubuntu on it, in a huge partition sda1 formatted with ext4 (no separate home), with an extended partition sda2 containing sda5 swap. Gparted within Salix live recognises none of this, seeing just an unformatted disk the size of the whole thing.

The persistence sees sda1 - but doesn't seem to work at all. From inside Ubuntu, I can see the persistence file, which is 1GB as I chose. I then looked at Gparted because my first hunch is maybe Salix can't read ext4 yet? But in that case, how did it save the persistence file?

My next step is to try a different machine, for instance the one downstairs that has Salix on it already (and almost Caffeine). But if there *is* a problem with ext4, it might be helpful to others to solve it. I imagine the most common scenario is ntfs.

A separate point, and just a minor gripe: the persistence setup program asks you to select a mount point. In my case, there was only the one, and it wasn't obvious that you needed to highlight it. I didn't, and clicked OK, and the window just closed. That meant two reboots before I realised it hadn't worked. Would it be possible to make it refuse to let you click on OK till you have selected something, or produce an error message, or maybe just change the wording slightly? I know it already says something like "select", but if you have one option and a button that says OK, the tendency is to just click :oops:

UPDATE The machine with Salix installed on it seems to have the same problem with persistence for Salix Live 13.1.1, i.e. it doesn't work. Interestingly, the login screen is skipped on this machine.

I don't think this is a corrupt download. The md5sums check out, and also because I was having some problems with burning I downloaded from a mirror as well as torrent, and the two files are identical (using diff). Also, everything else works.
User avatar
JRD
Salix Warrior
Posts: 950
Joined: 7. Jun 2009, 22:52
Location: Lyon, France

Re: problem with persistence

Post by JRD »

Hi !
mimosa wrote:Gparted within Salix live recognises none of this, seeing just an unformatted disk the size of the whole thing.
Hum very weird. Is your partitions mounted in /mnt on Salix Live ?
Could you post the output of the "mount" command when run on Salix Live on that machine please ?
Do you have some RAID configuration maybe ?
mimosa wrote:Would it be possible to make it refuse to let you click on OK till you have selected something, or produce an error message, or maybe just change the wording slightly?
Yes, I think it's possible.

Could you also post the output of "mount" on Salix Live with the persistence file on the downstairs machine please ?
Image
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

Re: problem with persistence

Post by mimosa »

Thanks JRD, I'll get back to you when I have a chance to access both machines.

Meanwhile, I just wonder if I'm missing something incredibly obvious such as - you have to tell Salix to save changes to its persistence file (it's not like this on Puppy for instance, though there sometimes seems to be an *option* not to save configuration changes). My test was to change the keyboard layout to Btazilian, and create a dummy folder on the desktop. The first time, I also did a lot of updates and changed settings in Firefox, but I didn't realise it was a testing scenario at that point.

It does seem rather odd that I should be having the same problem on two different machines that are also quite different in hardware terms. However, the failure to recognise this machine's partitions, even in Gparted, looks like a separate issue. And yet, the persistence wizard correctly reflects free space on the drive, which went down from ~33GB to ~32GB when I asked it to create the persistence file!
User avatar
JRD
Salix Warrior
Posts: 950
Joined: 7. Jun 2009, 22:52
Location: Lyon, France

Re: problem with persistence

Post by JRD »

mimosa wrote:Meanwhile, I just wonder if I'm missing something incredibly obvious such as - you have to tell Salix to save changes to its persistence file
You don't have to say anything.
But the scenario :
Boot into Salix Live, install persistence, make some changes, reboot
Will lead to "no changes have been stored".

This is explain in the text of the persistence wizard : after having install the persistence file, you need to reboot, and thereafter, any changes that you made will be saved, not before rebooting
This scenario works :
Boot into SAlix Live, install persistence, reboot, make some changes, reboot
Changes have been saved.

Or maybe I just explain this but you already got that.
Image
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

Re: problem with persistence

Post by mimosa »

Hi JRD,

Here is the output of mount from the machine with Salix already on it, but running Salix Live 13.1.1

Code: Select all

one[~]$ mount
rootfs on /mnt/live/ type rootfs (rw)
/dev/root on /mnt/live/ type ext2 (rw,relatime,errors=continue)
devtmpfs on /mnt/live/dev type devtmpfs (rw,relatime,size=506812k,nr_inodes=126703,mode=755)
/dev/loop1 on /mnt/live/lib/modules/2.6.33.4-smp/kernel/drivers type squashfs (ro,noatime)
/dev/sda2 on /mnt/live/mnt/sda2 type vfat (rw,noatime,fmask=0000,dmask=0000,allow_utime=0022,codepage=cp437,iocharset=iso8859-1,shortname=winnt,check=s,quiet,utf8,errors=remount-ro)
tmpfs on /mnt/live/memory type tmpfs (rw,relatime,size=609960k)
aufs on / type aufs (rw,relatime,si=2262466c,nowarn_perm)
/dev/loop2 on /mnt/live/memory/images/01-core.lzm type squashfs (ro,noatime)
/dev/loop3 on /mnt/live/memory/images/02-basic.lzm type squashfs (ro,noatime)
/dev/loop4 on /mnt/live/memory/images/03-full.lzm type squashfs (ro,noatime)
/dev/loop5 on /mnt/live/memory/images/04-common.lzm type squashfs (ro,noatime)
/dev/loop6 on /mnt/live/memory/images/05-kernel.lzm type squashfs (ro,noatime)
/dev/loop7 on /mnt/live/memory/images/06-livetools.lzm type squashfs (ro,noatime)
/dev/loop8 on /mnt/live/memory/images/07-live.lzm type squashfs (ro,noatime)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
usbfs on /proc/bus/usb type usbfs (rw)
tmpfs on /dev/shm type tmpfs (rw)
/dev/sr0 on /mnt/sr0 type iso9660 (ro,noatime)
/dev/sda7 on /mnt/sda7 type reiserfs (rw,noatime)
/dev/sda6 on /mnt/sda6 type ext3 (rw,noatime)
/dev/sda2 on /mnt/sda2 type vfat (rw,noatime,quiet,umask=0,check=s,shortname=winnt,utf8)
/dev/sda1 on /mnt/sda1 type fuseblk (rw,noatime,allow_other,blksize=4096)
gvfs-fuse-daemon on /home/one/.gvfs type fuse.gvfs-fuse-daemon (rw,nosuid,nodev,user=one)
one[~]$ 
The persistence file is there in sda2, the Windows data partition, and I can see it from within Salix Live.

I'll post from the other machine (with Ubuntu on it) in a while.

By the way, yes, I realised you need to reboot again for persistence to take effect. I've made some changes again to be double sure.

Once again, the login screen was skipped on this machine; this hasn't been happening upstairs.
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

Re: problem with persistence

Post by mimosa »

OK, here is the output from upstairs (single-boot Ubuntu machine)

Code: Select all

one[~]$ mount
rootfs on /mnt/live/ type rootfs (rw)
/dev/root on /mnt/live/ type ext2 (rw,relatime,errors=continue)
devtmpfs on /mnt/live/dev type devtmpfs (rw,relatime,size=380420k,nr_inodes=95105,mode=755)
/dev/loop1 on /mnt/live/lib/modules/2.6.33.4-smp/kernel/drivers type squashfs (ro,noatime)
/dev/sda1 on /mnt/live/mnt/sda1 type ext4 (rw,noatime,barrier=1,data=ordered)
tmpfs on /mnt/live/memory type tmpfs (rw,relatime,size=458288k)
aufs on / type aufs (rw,relatime,si=5607a722,nowarn_perm)
/dev/loop2 on /mnt/live/memory/images/01-core.lzm type squashfs (ro,noatime)
/dev/loop3 on /mnt/live/memory/images/02-basic.lzm type squashfs (ro,noatime)
/dev/loop4 on /mnt/live/memory/images/03-full.lzm type squashfs (ro,noatime)
/dev/loop5 on /mnt/live/memory/images/04-common.lzm type squashfs (ro,noatime)
/dev/loop6 on /mnt/live/memory/images/05-kernel.lzm type squashfs (ro,noatime)
/dev/loop7 on /mnt/live/memory/images/06-livetools.lzm type squashfs (ro,noatime)
/dev/loop8 on /mnt/live/memory/images/07-live.lzm type squashfs (ro,noatime)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
usbfs on /proc/bus/usb type usbfs (rw)
tmpfs on /dev/shm type tmpfs (rw)
/dev/sr0 on /mnt/sr0 type iso9660 (ro,noatime)
/dev/sda1 on /mnt/sda1 type ext4 (rw,noatime)
gvfs-fuse-daemon on /home/one/.gvfs type fuse.gvfs-fuse-daemon (rw,nosuid,nodev,user=one)
one[~]$ 
GParted is now showing the partitions correctly, so I`m not sure what was going on there.

During the boot process, after a line identifying the location of the persistent file at - from memory - /mnt/sda1/whatever the file is called, the next line says something about no changes or not possible to save so using memory

I thought I`d already done this, and perhaps I had, but I`m going to make an image file from the CD I`m using so as to be quite sure its isn`t a bad disk. It`s just quite hard to think of a plausible explanation for this happening on two completely different machines.
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

Re: problem with persistence

Post by mimosa »

Hi JRD,

The CD's md5sum checks out:

Code: Select all

one[~]$ dd if=/dev/cdrom bs=1024 count=683024 | md5sum
683024+0 records in
683024+0 records out
699416576 bytes (699 MB) copied, 141.454 s, 4.9 MB/s
f09431669caabef402073ec4814f34ad  -
one[~]$ su
Password: 
root[one]# eject
root[one]# exit
exit
one[~]$ 
one[~]$ dd if=/dev/cdrom bs=1024 count=683024 | md5sum
683024+0 records in
683024+0 records out
699416576 bytes (699 MB) copied, 135.413 s, 5.2 MB/s
f09431669caabef402073ec4814f34ad  -
one[~]$ 
I burned another one earlier, but both seem to be ok.

The persistence wizard on this (downstairs, Salix/Windows) machine is transposing the partition names and the sizes (ie available space on that partition). I chose sda2 and it said it had succesfully created the file on [some string of digits]. When I looked, I couldn't find the file on any partition. This didn't happen before, and so I still suspect the disk, even though md5sum checks out. Has anyone else been having any trouble of this kind?

I know there are other commonly-used checksums apart from md5sum. Could you send me one as a cross-check?
User avatar
mimosa
Salix Warrior
Posts: 3311
Joined: 25. May 2010, 17:02
Contact:

Re: problem with persistence

Post by mimosa »

On reflection, the disk must be ok, because I downloaded the image twice from different sources and both files are identical as well as having the correct md5sum. But then, why aren't other people having similar problems?
Post Reply