Create persistent USB live system for Mint 14 KDE

Questions about Grub, UEFI,the liveCD and the installer
Forum rules
Before you post read how to get help. Topics in this forum are automatically closed 6 months after creation.
Locked
Chaos99

Create persistent USB live system for Mint 14 KDE

Post by Chaos99 »

I'm trying (and failing) to create an USB live system from the Mint 14 KDE 64bit iso.

I've copied the iso to a fat32 partition using UNetbootin (from within Windows) checking the 'persistent' option and reserving 1gig of memory.
After that I used gparted (from another Mint installation) to create a second partition (ext2) called casper-rw and deleted the casper-rw file on the first partition.

The system boots up perfectly, but the second partition is not mounted and there is no persistence.

I've verified that UNetbooting added the 'persistent' option to the syslinux.cfg file.

I've tried Mint 14 Debian Edition before following the same procedure, just that the second partition had to be called live-rw. (The persistence worked, but I wasn't able to upgrade to the newest kernel from that half-year old iso.) I tried to rename the second partition like that also for the KDE version, but that failed too.

Has anybody successfully created a persistent live installation from Mint 14 KDE using a second partition and can help?

Thanks!
Last edited by LockBot on Wed Dec 28, 2022 7:16 am, edited 1 time in total.
Reason: Topic automatically closed 6 months after creation. New replies are no longer allowed.
usbtux

Re: Create persistent USB live system for Mint 14 KDE

Post by usbtux »

Persistence Partition

You need 2 usb drives or at least a live dvd and the usb drive.

From the live dvd

Using gparted -
Make a New Partition Table on the usb stick to remove all partitions.
Make 2 fat32 partitions 1st for live cd/usb
2nd for data
mark first partition as boot (under flags)

From windows
Install liveusb with unetbootin to 1st partition.

When using Unetbootin add 100mb of persistence.
Once Unetbootin has finished delete the FILE casper-rw.

From the live dvd
using gparted change the 2nd partition label to casper-rw and format to Ext2

Boot from flash - Persistence now active.

This seems very similar to what you've done- it should work.
Chaos99

Re: Create persistent USB live system for Mint 14 KDE

Post by Chaos99 »

That was one of the guides I used for my experiment.
I've re-done all the steps (replacing the Live CD with a Mint 13 installation on another machine).
I have a persistent system now, even if the casper-cw partition isn't listed in my mtab (as it was in the LMDE version). But if I mount it and take a look, the changed files indeed appear on that partition.

The downside: if I change only so much as adding a new user, the system doesn't start up anymore. ( This also worked flawlessly on the Debian Edition.)

Have you actually tried with a Mint 14 KDE?

Thanks anyway...
ada101

Re: Create persistent USB live system for Mint 14 KDE

Post by ada101 »

Thanks for this, working great! In fact, I am writing this post from the live os now!! I did something different to what you said. Instead of making the 2 partitions as the first two partitions, I made an extended partition and put them both there. the reason for this is because i need a fat32 partition at the start of the drive.. Look:
http://i45.tinypic.com/ylzdw.png
:)
brainout

Re: Create persistent USB live system for Mint 14 KDE

Post by brainout »

ada101 wrote:Thanks for this, working great! In fact, I am writing this post from the live os now!! I did something different to what you said. Instead of making the 2 partitions as the first two partitions, I made an extended partition and put them both there. the reason for this is because i need a fat32 partition at the start of the drive.. Look:
http://i45.tinypic.com/ylzdw.png
:)
How did you move the boot partition into the extended partition? Did you instead really copy it, and then relabel? Thank you for your time in reply! I'm so excited about your idea, I made a video about it, here: http://www.youtube.com/watch?v=pUR5hu2O_6s I think I solved it, but wanted to know your own steps, if you are willing to share them.

Why this question: I sooooo want to be able to get Linux on a stick, especially Mint 13. The value of doing so is enormous. Everyone in Windows will want to buy these, so Linux could make money while preserving Open Source, if the distro for 13 were available on a stick. No shame in charging for it. The few hackers can still do their hacking, but the developers got paid upfront. Perfectly moral thing, to charge for Linux Mint 13 on a stick. For on a stick, one can use Linux to fix Windows crashes and do a bunch of other stuff on almost any computer, without having to go the wubi route, without having to 'touch' the computer used. So not merely for 'trying', but for 'buying'. Computer in your pocket, better than Raspberry Pi. Potentially. If on a stick with PERSISTENCE equal to the full capacity of the stick. Just think: you take everything with you, so if you wanted to go to a tablet which can boot in Linux, well just plug it in. Want to run your netbook as you travel? Just plug it in. How about an Ultrabook, laptop, borrowed computer in the airport lounge? Plug it in. Whole new dimension, of mobile computing.

Linux Mint could justifiably sell the sticks (preferably SanCruzer, Kingston) at 64GB, pre-installed with Mint 13 plus extra disk management tools, for what.. $150 each? I'd buy sticks thus configured, if they worked out of the box. Many businesses would. Huge market of small businesses like mine, mom-and-pop consultants (or doctors, lawyers, brokers, accountants, tradesmen) -- who don't have IT departments, but make good money.

Although the sticks could be cloned, it's better to have several. Say, selling them in packs of 2 or three, or with bundle discounts. A consultant's time is worth over $100 per hour minimum, so the hour it takes to clone* is already nearly the value of purchase. But purchase, takes five minutes; and during the rest of the hour he can make money to pay for the stick.

Example: I have 24 computers which only I own, only I use, no one else lives with me and no one is allowed in. Confidential data. Okay, so now I want to do stuff to the 12 computers which have the specs Mint 13 needs to run. So I put the stick into each one, do my disk management, and.. done! I don't have to do 12 installations. I don't have to set up a network. It's easier to just walk to each machine. Some of the Linux software runs best on one machine, some of the other software runs better on another. So, just walk to the machine which best works with that software.

Granted, the example is kinda rare, but its inherent pattern can apply to a number of situations. A truly live distro on a stick is valuable. Far better than CD, if PERSISTENCE is had. So far, the only distros which end up having persistence are Fedora 17 and Mint 13, but each seem capped at 1 GB above their own software, rather than addressing the whole of the capacity of the stick. That's why your solution is so intriguing. I used Unetbootin, and no matter how I tweak it, I can't bypass that add-1GB barrier in Mint 13 or Fedora. Fedora has its own Live USB installer, but the limits post-creation, are the same. Since I can't move the /home folder and can't get the rest of the stick to store the data or programs outside /root (because I'm given no option from the package managers where to install the programs), this 1GB limit will soon fill up. I can partition the remaining space as NTFS and then it is separately addressable (it's not using the space when marked ext2,3,4); but again all the package installers only go to /root or /home. This is an unhappy limitation.

This persistence doesn't occur with the latest Linux distros with any predictability. It's hit and miss. Alas, Mint 14 -- which I can't yet make persistent -- has no good desktop customization and is glitchy, really tempermental about permissions (like Ubuntu 13.04 and Fedora 18) so it's not viable, not useful at all. But 13 is great, especially if one wants to wean off Windows.

*It takes an hour to clone, because the typical buyer is not interested in becoming a Linux geek, so has to review the rules again for cloning. Takes time to do that. Then has to wait, and since it takes 20 minutes to clone 35GB on a standard external hard drive, it probably takes an hour to do it on a flash drive. Even if less, again the average buyer is NOT a Linux geek so would have to spend time refreshing on the steps for cloning. Not likely, he'll want to clone, if he can just BUY another stick, instead. If I could avoid learning Linux altogether, I would, so I could spend the same amount of time serving my customers. Millions of people share my situation and needs.
Last edited by brainout on Sun Feb 17, 2013 8:16 pm, edited 1 time in total.
usbtux

Re: Create persistent USB live system for Mint 14 KDE

Post by usbtux »

brainout - your long post what are you trying to say??

Who the fudge would pay $150 for a live usb with persistence?

Anyhoo, Instead of a live usb with persistence, the other way would be to install to a usb stick. http://youtu.be/gVCfeFOxGLQ External install - usb stick/drive.

With the flash stick you may not want to use swap and set up a ram drive for firefox/chrome http://www.usbtux.hostzi.com/flash_install
brainout

Re: Create persistent USB live system for Mint 14 KDE

Post by brainout »

usbtux wrote:brainout - your long post what are you trying to say??

Who the fudge would pay $150 for a live usb with persistence?

Anyhoo, Instead of a live usb with persistence, the other way would be to install to a usb stick. http://youtu.be/gVCfeFOxGLQ External install - usb stick/drive.

With the flash stick you may not want to use swap and set up a ram drive for firefox/chrome http://www.usbtux.hostzi.com/flash_install
I would gladly pay $150 instead of having to spend a week without showers, combing the internet to read the contradictory instructions on how to make a stick. Ubuntu's and Fedora's own instructions, don't work. ALL videos in Youtube on the topic, don't work. Only the thread here, and specifically ada101's post, DOES work. So just think of all the time-money wasted on bad instructions, versus the saving post of ada101, which is based on YOUR post? You both saved me a ton of time, and I wish I'd seen YOUR posts before the others.

So as I type, I'm making a USB stick for Kubuntu with Unetbootin, following ada101's change, since her method helped me create a true persistence stick-wide. In short, your instructions are the base, and her changes made it work without error. So when I need to make another stick, I can switch between the two distros. For as you know, you can't use GParted on a stick with the same distro as the one running live. Gotta always have at least two distros on a stick. I have three (Fedora 17 is third, I erased Ubuntu 13.04, as it's too buggy). But the third needs to be redone. I also have Mint 13 on another stick which is NOT done per ada101's change, but rather per the written instructions from Unetbootin. Those instructions only grant 1 GB persistence above the installed OS, so there is no room to add packages. So I'll have to redo it the way she did it, to get full access. Am in the middle of that now (well, after a video stops rendering).

Even when done, look at the time cost once a person is savvy: 15-30 minutes to refamiliarize with the procedure, another 15 to use GParted, another 30 to download the iso, another 15 to make the stick in Unetbootin, another 15 to get rid of the casper-rw file. It's CHEAPER to buy the stick, than to make it. I could have finished a client tax return in the same amount of time, and charged the client $1,000 he'd happily pay.

Big market for this. Not everyone so undervalues Linux that they insist on 'free' software. To the business user, 'free' means FREEDOM, not no cost. And, it requires Open Source for tweaking, but almost no one in business would do his own tweaking, but would rather HIRE it.

Enter a stick for $150 to save time and money from Windows crashes, and 'funding' will never be a problem, again. When a Windows machine dies, the stick can be used to clone, backup, correct, rescue. When someone wants to use some feature unique to Linux (like, partitioning a pendrive), he can just plug in the drive. Far easier than using CDs. When someone wants to integrate Windows and Linux features, he just plugs the stick into the Windows computer and has access (assuming the Windows machine allows it, as one's OWN machines, will). Many uses. Mind you, I prefer Windows, still. It's smoother. But post-XP, Windows is rapidly becoming less customizable and more paranoid about permissions, which makes it far less useful. Sadly, Mint 14, Ubuntu 12.10 and after, Fedora 18 share those same disadvantages. But earlier versions still work.

Someday someone will figure out that there is a large customer base who wants CONFIGURABILITY and will gladly pay for it. They don't want to be hackers. They don't want to alter the source code themselves. They want someone else to do it, and will pay; but they don't want to be slammed down every five seconds with permissions issues or lack of customization. Sadly, right now Linux is moving in the same bad direction as Windows. So that is the ultimate reason to need an older version of Linux on a stick: to restore one's own FREEDOM to use one's own machine. That's how I got into Linux, and that's why I'll stay.
ada101

Re: Create persistent USB live system for Mint 14 KDE

Post by ada101 »

I'm sorry for the long reply, didn't have notifications enabled for this thread :( . I will post a video tutorial within the next week depending on time :) . Look forward to posting it :)

The post i'm replying to:
brainout wrote:
ada101 wrote:Thanks for this, working great! In fact, I am writing this post from the live os now!! I did something different to what you said. Instead of making the 2 partitions as the first two partitions, I made an extended partition and put them both there. the reason for this is because i need a fat32 partition at the start of the drive.. Look:
http://i45.tinypic.com/ylzdw.png
:)
How did you move the boot partition into the extended partition? Did you instead really copy it, and then relabel? Thank you for your time in reply! I'm so excited about your idea, I made a video about it, here: http://www.youtube.com/watch?v=pUR5hu2O_6s I think I solved it, but wanted to know your own steps, if you are willing to share them.

Why this question: I sooooo want to be able to get Linux on a stick, especially Mint 13. The value of doing so is enormous. Everyone in Windows will want to buy these, so Linux could make money while preserving Open Source, if the distro for 13 were available on a stick. No shame in charging for it. The few hackers can still do their hacking, but the developers got paid upfront. Perfectly moral thing, to charge for Linux Mint 13 on a stick. For on a stick, one can use Linux to fix Windows crashes and do a bunch of other stuff on almost any computer, without having to go the wubi route, without having to 'touch' the computer used. So not merely for 'trying', but for 'buying'. Computer in your pocket, better than Raspberry Pi. Potentially. If on a stick with PERSISTENCE equal to the full capacity of the stick. Just think: you take everything with you, so if you wanted to go to a tablet which can boot in Linux, well just plug it in. Want to run your netbook as you travel? Just plug it in. How about an Ultrabook, laptop, borrowed computer in the airport lounge? Plug it in. Whole new dimension, of mobile computing.

Linux Mint could justifiably sell the sticks (preferably SanCruzer, Kingston) at 64GB, pre-installed with Mint 13 plus extra disk management tools, for what.. $150 each? I'd buy sticks thus configured, if they worked out of the box. Many businesses would. Huge market of small businesses like mine, mom-and-pop consultants (or doctors, lawyers, brokers, accountants, tradesmen) -- who don't have IT departments, but make good money.

Although the sticks could be cloned, it's better to have several. Say, selling them in packs of 2 or three, or with bundle discounts. A consultant's time is worth over $100 per hour minimum, so the hour it takes to clone* is already nearly the value of purchase. But purchase, takes five minutes; and during the rest of the hour he can make money to pay for the stick.

Example: I have 24 computers which only I own, only I use, no one else lives with me and no one is allowed in. Confidential data. Okay, so now I want to do stuff to the 12 computers which have the specs Mint 13 needs to run. So I put the stick into each one, do my disk management, and.. done! I don't have to do 12 installations. I don't have to set up a network. It's easier to just walk to each machine. Some of the Linux software runs best on one machine, some of the other software runs better on another. So, just walk to the machine which best works with that software.

Granted, the example is kinda rare, but its inherent pattern can apply to a number of situations. A truly live distro on a stick is valuable. Far better than CD, if PERSISTENCE is had. So far, the only distros which end up having persistence are Fedora 17 and Mint 13, but each seem capped at 1 GB above their own software, rather than addressing the whole of the capacity of the stick. That's why your solution is so intriguing. I used Unetbootin, and no matter how I tweak it, I can't bypass that add-1GB barrier in Mint 13 or Fedora. Fedora has its own Live USB installer, but the limits post-creation, are the same. Since I can't move the /home folder and can't get the rest of the stick to store the data or programs outside /root (because I'm given no option from the package managers where to install the programs), this 1GB limit will soon fill up. I can partition the remaining space as NTFS and then it is separately addressable (it's not using the space when marked ext2,3,4); but again all the package installers only go to /root or /home. This is an unhappy limitation.

This persistence doesn't occur with the latest Linux distros with any predictability. It's hit and miss. Alas, Mint 14 -- which I can't yet make persistent -- has no good desktop customization and is glitchy, really tempermental about permissions (like Ubuntu 13.04 and Fedora 18) so it's not viable, not useful at all. But 13 is great, especially if one wants to wean off Windows.

*It takes an hour to clone, because the typical buyer is not interested in becoming a Linux geek, so has to review the rules again for cloning. Takes time to do that. Then has to wait, and since it takes 20 minutes to clone 35GB on a standard external hard drive, it probably takes an hour to do it on a flash drive. Even if less, again the average buyer is NOT a Linux geek so would have to spend time refreshing on the steps for cloning. Not likely, he'll want to clone, if he can just BUY another stick, instead. If I could avoid learning Linux altogether, I would, so I could spend the same amount of time serving my customers. Millions of people share my situation and needs.
User avatar
catweazel
Level 19
Level 19
Posts: 9763
Joined: Fri Oct 12, 2012 9:44 pm
Location: Australian Antarctic Territory

Re: Create persistent USB live system for Mint 14 KDE

Post by catweazel »

usbtux wrote: Once Unetbootin has finished delete the FILE casper-rw.
[...]
From the live dvd
using gparted change the 2nd partition label to casper-rw and format to Ext2
Interesting. Does that speed up the live OS?

I'm asking because I created a Mint KDE live cd using unetbootin with persistence and it was slower than a week full of wet Sundays.
"There is, ultimately, only one truth -- cogito, ergo sum -- everything else is an assumption." - Me, my swansong.
usbtux

Re: Create persistent USB live system for Mint 14 KDE

Post by usbtux »

TehGhodTrole wrote:
usbtux wrote: Once Unetbootin has finished delete the FILE casper-rw.
[...]
From the live dvd
using gparted change the 2nd partition label to casper-rw and format to Ext2
Interesting. Does that speed up the live OS?

I'm asking because I created a Mint KDE live cd using unetbootin with persistence and it was slower than a week full of wet Sundays.
Liveusb +persistence /installed to usb are generally slower because of the read write speeds in comparison to a install on hdd, even with a fast/good flash drive. Add in the fact that its writing to a virtual filesystem in memory it only slows them down.
usbtux

Re: Create persistent USB live system for Mint 14 KDE

Post by usbtux »

TehGhodTrole wrote:
usbtux wrote: Once Unetbootin has finished delete the FILE casper-rw.
[...]
From the live dvd
using gparted change the 2nd partition label to casper-rw and format to Ext2
No doesn't speed it up but allows you to get over a size limit for the persistence file (4G). Saying that though it should speed it up as its not writing to a virtual system.
littlebigman

Re: Create persistent USB live system for Mint 14 KDE

Post by littlebigman »

Hello

Since this thread is from 2013, I wanted to check in case things changed between then and now (Aug 2014).

Using Rufus, a friend successfully burned a USB keydrive with the 64-bit Mint Cinnamon ISO file.

However, Rufus doesn't support persistence* so that my friend loses all the changes made while using Mint (ie. choice of a non-US keyboard layout, new bookmarks in Firefox, etc.)

Should he use another Windows application** to burn the ISO and make the USB persistent, or is there a feature in Mint that can write changes to the USB keydrive?

Thank you.

* "Do you plan to add Linux persistent data support (e.g. for Ubuntu)? No (unless I get a lot of free time on my hands, which is unlikely to happen).
Once you boot the OS, you should have everything you need to create some space for peristent data, so there's no real need for Rufus to do that."

** I know of Unetbootin, Pendrive Linux Universal USB Installer, and LinuxLive USB Creator
Locked

Return to “Installation & Boot”