rolisoft / wsl-distribution-switcher Goto Github PK

Scripts to replace the distribution behind Windows Subsystem for Linux with any other Linux distribution published on Docker Hub.

License: MIT License

Python 74.83% Shell 14.04% C 11.12%

wsl-distribution-switcher's Issues

Couldn't get root access in Fedora

I used the script to install fedora and enabled the hook script but still can not get access to the root account neither can install sudo
note: I already put ROOTPASSWD in system environment variables - System Variables

also how can I delete installed image as I tried to re install fedora after removing the sample from the hook file and now the switch script show I have two copies of fedora:latest

Windows error 139

Hey, trying to set up Centos 6.6, no real errors in console while installing but bash keeps crashing with error 139.

C:\dev\WSL-Distribution-Switcher>get-source centos:6.6
[*] Fetching official-images info for centos:6.6...
[*] Fetching Dockerfile from repo CentOS/sig-cloud-instance-images/8911843d9a6cc71aadd81e491f94618aded94f30/docker...
[*] Downloading archive https://raw.githubusercontent.com/CentOS/sig-cloud-instance-images/8911843d9a6cc71aadd81e491f94618aded94f30/docker/centos-6.6-20150304_1234-docker.tar.xz...
[*] Rootfs archive for centos:6.6 saved to rootfs_centos_6.6.tar.xz.

C:\dev\WSL-Distribution-Switcher>install centos:6.6
[*] Probing the Linux subsystem...
[*] Default user is tost at /home/tost.
[*] Reading /etc/{passwd,shadow,group,gshadow} entries for root and tost...
[*] Beginning extraction...
[*] Backing up current rootfs to rootfs_ubuntu_trusty...
[*] Switching to new rootfs...
[*] Writing entries of root and tost to /etc/{passwd,shadow,group,gshadow}...

C:\dev\WSL-Distribution-Switcher>bash

C:\dev\WSL-Distribution-Switcher>echo %errorlevel%
139

From windows docs:

ERROR_SUBST_TO_SUBST
139 (0x8B)
The system tried to substitute a drive to a directory on a substituted drive.

Any idea what's causing this?
Awesome project btw :))

edit

Tested on another PC, same issue, Centos 7 works fine. Weird...

Removing a rootfs

The following distributions are currently installed:

  - debian:latest*
  - debian:latest
  - fedora:latest
  - ubuntu:trusty

To switch back to the default distribution, specify ubuntu:trusty as the argument.

I have to versions of debian:lates installed. How do I remove one of them ?

Install process incorrectly extracts tarball in case-retentive fashion

After manually fetching Gentoo's stage3-amd64-nomultilib-20170525.tar.bz2 and using install.exe from v0.1.0 to install it under Windows 10 (1703), I discovered that the tarball was not extracted properly:-

$ cd /usr/share/terminfo/
$ echo *
1 2 3 4 5 6 7 8 9 a b c d e f g h i j k L M n o P q r s t u v w X z

This is missing eight additional directories. Had it been extracted properly, the list would have appeared as follows:-

$ echo *
1 2 3 4 5 6 7 8 9 a A b c d e E f g h i j k l L m M n N o p P q Q r s t u v w x X z

The damage is not limited to files owned by ncurses. However, this particular example is interesting because it manifests as weird terminal behaviour. This is because Windows sets TERM=xterm-256color, but the corresponding terminfo file cannot be located. Normally it would be beneath the lower-case x directory but, instead, it is beneath the upper-case X directory.

I recall previously using an older version of WSL-Distribution-Switcher with a standalone Python, without issue.

SOLVED: sudo: unable to create sockets: Socket type not supported (CentOS on WSL)

Hello,

@jdmansour pointed out in Atlas on Windows that he was also getting sudo: unable to create sockets: Socket type not supported. This was the only reference to sudo: unable to create sockets: Socket type not supported that I was able to find.

Since sudo works under Ubuntu but fails under CentOS, I thought it must be a CentOS issue rather than a WSL issue.

If you're running as root, you can use sudo -V to dump a detailed sudo configuration, along with ./configure compile time flags. After a lot of trial and error, I was able to get sudo working under CentOS via WSL Distribution Switcher.

First, download the sudo sourcecode from sudo.ws.
Then you'll need to install a boatload of dependancies to get this to build.

yum install file make zlib zlib-devel \
gcc vim svn wget make \
sssd sssd-libwbclient-devel \
pam-devel libselinux-devel

Then you'll want to configure with:

./configure  -v --with-all-insults --with-pam \
--with-nsswitch=no --without-ldap --without-sendmail \
--with-logfac=authpriv --with-logging=syslog \
--with-editor=/bin/vi --with-env-editor \
--with-timeout=15 --with-password-timeout=0 \
--with-passprompt="[sudo] password for %p:" \
--without-lecture --with-tty-tickets \
--disable-root-mailer --enable-admin-flag \
--with-sssd --with-sssd-lib=/usr/lib64 \
--with-selinux --enable-zlib \
--without-fqdn

Last, you'll want to edit /etc/sudoers. On the very last line, you'll want to put:

me   ALL=(ALL) NOPASSWD: ALL

Of course me should be your actual username. You'll also want to make sure to include a blank line after this line, as it also seems to be required by sudo. This should fix: sudo: no tty present and no askpass program specified errors.

I'm sure that it's not the best solution, but it's a viable way to get sudo working on CentOS under WSL. I wish I could think of a way to include this into the post_install script but it's such a long winded process, building a custom RPM would seem to be the most viable option, but then is WSL-Distribution-Switcher going to want to host packages that handle compatability for certain programs?

Remove the chroot hack for 14936

Hiya @RoliSoft ; New build (14936) adds support for chroot, so you can probably fix your arch installer hack (if you want to check for WinBuild number).

Cheers!

Clean Arch Install, latest pull

Heya, @RoliSoft

I have a log of a brand new install off a new arch.SFS and you mislocated the libmockchroot in the preinstall script, probably forgot to change its location later on.

Also, lots of problems with sigchecks failing even after fixing that on my end. It says it can't look up the keys remotely.

Edit: Oops, I pulled it like 4 hours ago =.=.

Edit2: Nope, I was right, you are overwriting it after all. Find where you write "libfakechroot" and you'll see you're overwriting it.

Anyway, after fixing it, I think you need to add two lines to the postinstall script for arch:
pacman-key --init

pacman-key --populate archlinux

Edit: Maybe I'll just do a pull request

Anyway, afterwards, I'm still getting this acl error afterwards

Edit 3: I'm going to fix this properly then submit a proper pull request. The last one was messed up.

debian doesn't install sudo

i just installed debian, but sudo didn't get installed. i set the regular username/password, but how do i su into root to install sudo? thanks

Failed to run hook in WSL when using Fedora:latest

Is this a know issue? I am not finding anyone who documented such issue:

[] Installing tools...
error: rpmdb: BDB0113 Thread/process 7/140163040478976 failed: BDB1507 Thread died in Berkeley DB library
error: db5 error(-30973) from dbenv->failchk: BDB0087 DB_RUNRECOVERY: Fatal error, run database recovery
error: cannot open Packages index using db5 - (-30973)
error: cannot open Packages database in /var/lib/rpm
Error: Error: rpmdb open failed
[!] Failed to run hook in WSL: Command '['cmd', '/C', 'C:\windows\System32\bash.exe', '-c', 'REGULARUSER="ikky" WINVER="14393" /root/hook_postinstall_all.sh']' returned non-zero exit status 1.
[] Switching default user back to ikky...

Store the switch.py rootfs files somewhere else or remove their /etc/shadow files for security reasons.

Hiya Roli, I found a kind of security flaw in the design of the distro switcher.

Suppose you have a no-goodnik on your bash install through ssh. Right now, if you've used the WSL-switcher, they can just go do

cd /mnt/c/Users/Username/AppData/Local/lxss

then if there is any other rootfs than the default one, say rootfs_ubuntu_xenial, they can just enter that with cd and read off of /etc/shadow.

Two possible solutions: One, regenerate/delete the /etc/passwd and /etc/shadow files in switch.py, or store the rootfs_distro_tag folders inside of %localappdata%\lxss\rootfs\sbin, since Microsoft protects viewing that from bash, and bash prevents non-root users from viewing it (because of the nature of sbin).

Different path for lxrun and bash

Hi, on my system lxrun and bash are not in c:\windows\sysnative but in c:\windows\system32

I fixed the scripts by removing the c:\windows\sysnative and allow the scripts to run the the lxrun and bash from PATH, but that may conflict with users using cygwin/msys2.

An option to select the path for lxrun/bash should be provided or to verify the executables before.

Thanks!

cpio: cap_set_file failed - Operation not supported

Thanks to Dist Switch I've changed my WSL from Ubuntu to CentOS.

I got my sudo problems worked out, but there are a handful of packages that I can't install due to the following error: error: unpacking of archive failed on file /some/file/name: cpio: cap_set_file failed - Operation not supported

Searching for this error, I found several references to Docker users having the same problem, but their solution seems to involve the Docker client/server and doesn't seem to apply back to WSL's VolFs filesystem.

Obviously the ideal solution would be for WSL to support every possible operation, but the fact that this mostly works on Ubuntu but fails to CentOS makes me wonder about other options. Unfortunately, I'm not terribly familiar with cpio and not really sure where to go next.

CentOS packages I've seen impacted are httpd and iputils

For example, try sudo yum install -y httpd
···
Loaded plugins: ovl, verify
Resolving Dependencies
--> Running transaction check
---> Package httpd.x86_64 0:2.4.6-40.el7.centos.4 will be installed
--> Finished Dependency Resolution

Dependencies Resolved

======================================================================================================================================================== Package Arch Version Repository Size ========================================================================================================================================================Installing:
httpd x86_64 2.4.6-40.el7.centos.4 updates 2.7 M

Transaction Summary
========================================================================================================================================================Install 1 Package

Total download size: 2.7 M
Installed size: 9.4 M
Downloading packages:
httpd-2.4.6-40.el7.centos.4.x86_64.rpm | 2.7 MB 00:00:02
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction
Installing : httpd-2.4.6-40.el7.centos.4.x86_64 1/1 Error unpacking rpm package httpd-2.4.6-40.el7.centos.4.x86_64
error: unpacking of archive failed on file /usr/sbin/suexec: cpio: cap_set_file
Verifying : httpd-2.4.6-40.el7.centos.4.x86_64 1/1

Failed:
httpd.x86_64 0:2.4.6-40.el7.centos.4

Complete!
···

When iputils fails, it's on:

Error unpacking rpm package iputils-20121221-7.el7.x86_64
error: unpacking of archive failed on file /usr/bin/ping: cpio: cap_set_file

Can't install CentOS 6

I'm trying to install CentOS 6. Fetching the filesystem works, but I get an error message during installation (see below). It is the same error message that you got here under "Installing the new rootfs".

If I try to start WSL with bash.exe afterwards, nothing happens. No error message, it just returns me to the prompt.

The issue seems to be that the old CentOS version's adduser doesn't take the --quiet parameter, and that trips up lxrun.exe. However /etc/passwd and /etc/shadow seem to be edited correctly. Any ideas how to get this running?

> python get-source.py centos:6
> python install.py centos:6
[*] Probing the Linux subsystem...
[*] Default user is jason at /home/jason.
[*] Switching default user to root...
[*] Reading /etc/{passwd,shadow,group,gshadow} entries for root and jason...
[*] Copying rootfs_centos_6.tar.xz to /root/rootfs-temp...
[*] Beginning extraction...
[*] Waiting for the Linux subsystem to exit...
[*] Backing up current rootfs to rootfs_ubuntu_trusty...
[*] Switching to new rootfs...
[*] Writing entries of root and jason to /etc/{passwd,shadow,group,gshadow}...
[*] Switching default user back to jason...
/usr/sbin/adduser: unrecognized option '--quiet'
Usage: adduser [options] LOGIN
       adduser -D
       adduser -D [options]

Options:
  -b, --base-dir BASE_DIR       base directory for the home directory of the
                                new account
  -c, --comment COMMENT         GECOS field of the new account
  -d, --home-dir HOME_DIR       home directory of the new account
  -D, --defaults                print or change default useradd configuration
  -e, --expiredate EXPIRE_DATE  expiration date of the new account
  -f, --inactive INACTIVE       password inactivity period of the new account
  -g, --gid GROUP               name or ID of the primary group of the new
                                account
  -G, --groups GROUPS           list of supplementary groups of the new
                                account
  -h, --help                    display this help message and exit
  -k, --skel SKEL_DIR           use this alternative skeleton directory
  -K, --key KEY=VALUE           override /etc/login.defs defaults
  -l, --no-log-init             do not add the user to the lastlog and
                                faillog databases
  -m, --create-home             create the user's home directory
  -M, --no-create-home          do not create the user's home directory
  -N, --no-user-group           do not create a group with the same name as
                                the user
  -o, --non-unique              allow to create users with duplicate
                                (non-unique) UID
  -p, --password PASSWORD       encrypted password of the new account
  -r, --system                  create a system account
  -R, --root CHROOT_DIR         directory to chroot into
  -s, --shell SHELL             login shell of the new account
  -u, --uid UID                 user ID of the new account
  -U, --user-group              create a group with the same name as the user
  -Z, --selinux-user SEUSER     use a specific SEUSER for the SELinux user mapping

[!] Failed to switch default user in WSL: Command '['cmd', '/C', 'C:\\Windows\\System32\\lxrun.exe', '/setdefaultuser', 'jason']' returned non-zero exit status 4294967295
Error in atexit._run_exitfuncs:
Traceback (most recent call last):
  File "install.py", line 119, in switch_user_back
    subprocess.check_output(['cmd', '/C', lxpath + '\\lxrun.exe', '/setdefaultuser', user])
  File "C:\Users\jdman\AppData\Local\Programs\Python\Python35\lib\subprocess.py", line 626, in check_output
    **kwargs).stdout
  File "C:\Users\jdman\AppData\Local\Programs\Python\Python35\lib\subprocess.py", line 708, in run
    output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command '['cmd', '/C', 'C:\\Windows\\System32\\lxrun.exe', '/setdefaultuser', 'jason']' returned non-zero exit status 4294967295

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "install.py", line 123, in switch_user_back
    exit(-1)
  File "C:\Users\jdman\AppData\Local\Programs\Python\Python35\lib\_sitebuiltins.py", line 26, in __call__
    raise SystemExit(code)
SystemExit: -1

Install.py homedirw is broken somehow

Latest version, when using install.py, it fails to move the homedir to the new temp directory because "homedirw" gets translated as

%localappdir%\lxss\home/usernameblah

instead of %localappdir%\lxss\home\usernameblahblah

Not sure exactly why the stripping thing that you do doesn't work right, but it doesn't work properly right now.

wmic not checked to see if it is on the path early enough

Running the hook causes wmic to be invoked, which fails if C:\Windows\system32\wbem is not on the PATH. It would be nice if wmic was searched for earlier and an error message with the appropriate suggestion on how to fix the issue was done before actually installing the distribution so that it could be fixed and the install run again

Unable to run get-sources

Can't even start the thing. Output below. I don't know what other info I can provide.

c:\Users\Chris\Downloads>python get-source.py opensuse:tumbleweed
Traceback (most recent call last):
  File "get-source.py", line 10, in <module>
    ensure_ca_load()
  File "c:\Users\Chris\Downloads\utils.py", line 85, in ensure_ca_load
    if ssl.create_default_context().cert_store_stats()['x509_ca'] == 0:
AttributeError: 'module' object has no attribute 'create_default_context'

c:\Users\Chris\Downloads>python --version
Python 3.3.5

Linux From Scratch?

I would love to have an LFS system in my WSL. What would be the best way to get there? I can build it all under WSL, I think (been a while since I tried). Should I try to put it in a SquashFS archive and then unsquash it as you mention in your blog for ArchLinux?

Can't install ubuntu xenial

I am hitting this after a fresh re-install of WSL and have Ubuntu:trusty installed:

python get-source.py ubuntu:xenial
[] Fetching official-images info for ubuntu:xenial...
[] Fetching Dockerfile from repo tianon/docker-brew-ubuntu-core/1a5cb40f41ac4829d8c301ccd2cf3b7a13687a8b/xenial...
[] Downloading archive https://raw.githubusercontent.com/tianon/docker-brew-ubuntu-core/1a5cb40f41ac4829d8c301ccd2cf3b7a13687a8b/xenial/ubuntu-xenial-core-cloudimg-amd64-root.tar.gz...
[] Rootfs archive for ubuntu:xenial saved to rootfs_ubuntu_xenial.tar.gz.

python get-prebuilt.py ubuntu:xenial
[] Requesting authorization token...
[] Fetching manifest info for ubuntu:xenial...
[] Downloading layer sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4...
[] Downloading layer sha256:6d9ef359eaaa311860550b478790123c4b22a2eaede8f8f46691b0b4433c08cf...
[] Downloading layer sha256:9654c40e9079e3d5b271ec71f6d83f8ce80cfa6f09d9737fc6bfd4d2456fed3f...
[] Downloading layer sha256:e8db7bf7c39fab6fec91b1b61e3914f21e60233c9823dd57c60bc360191aaf0d...
[] Downloading layer sha256:f8b845f45a87dc7c095b15f3d9661e640ebc86f42cd8e8ab36674846472027f7...
[] Downloading layer sha256:d54efb8db41d4ac23d29469940ec92da94c9a6c2d9e26ec060bebad1d1b0e48d...
[*] Rootfs archive for ubuntu:xenial saved to rootfs_ubuntu_xenial.tar.gz.

python install.py ubuntu:xenial
[] Probing the Linux subsystem...
[] Default user is ikky at /home/ikky.
[] Reading /etc/{passwd,shadow,group,gshadow} entries for root and ikky...
[] Beginning extraction...
[!] Failed to extract archive: [WinError 126] The specified module could not be found

P.S: WSL was NOT running when I was executing these commands.

Alpine Linux

Excuse me, did anyone install Alpine Linux?
python install.py alpine:latest
... OK ...
But, Bash cant't run.

would you implement a gentoo rootfs

too young too simple,sometime navie

cannot install centos:latest

python install.py rootfs_centos_latest.tar.xz
[*] Probing the Linux subsystem...
[!] Failed to get home directory of default user in WSL: [WinError 2] 。

dnf on fedora 25 results in a broken rpm database

I don't know what the issue is, but installing fedora 25 and then doing a dnf update results in the rpm database being broken and trying to repair it failed.

I doubt that the issue has anything to do with wsl-distribution-switcher, but wanted to note if in case anyone else ran in to this or anyone has a workaround

Manjaro

Are there any plans to add support for Manjaro? It's at no.3 on DistroWatch, so it might make sense to support it if possible.

Failed to fetch manifest info for distro: HTTP Error 404: NOT FOUND

Not sure why but I seem to be getting errors for any of the tagged official distros, but not 3rd party. Anyone have any ideas why this might be?


C:\Users\USER\Code\WSL-Distribution-Switcher>python get-prebuilt.py alpine:latest
[*] Requesting authorization token...
[*] Fetching manifest info for alpine:latest...
[!] Failed to fetch manifest info for alpine: HTTP Error 404: NOT FOUND

C:\Users\USER\Code\WSL-Distribution-Switcher>python get-prebuilt.py ubuntu:trusty
[*] Requesting authorization token...
[*] Fetching manifest info for ubuntu:trusty...
[!] Failed to fetch manifest info for ubuntu: HTTP Error 404: NOT FOUND

C:\Users\USER\Code\WSL-Distribution-Switcher>python get-prebuilt.py pritunl/archlinux
[*] Requesting authorization token...
[*] Fetching manifest info for pritunl/archlinux:latest...
[*] Downloading layer sha256:fa15d8aec9382c4924fd3ec2f716d6d7733acdcb5ce2dff508e0a783dbe2d469...

switch to centos bash.exe is cant work

Hello!

Ask a question!
there is no problem when running the python program,
download and install doker rootfs! but bash.exe is not working.
run bash.exe directly after the exit.he will not enter the bash command line mode.

when i switch back to the default subsystem bash.exe is running

Windows version 10.0.14393

how should I fix it？

-----command log-----

C:\WSL-Distribution-Switcher>install.py centos:6.6
[*] Probing the Linux subsystem...
[*] Default user is root at /root.
[*] Reading /etc/{passwd,shadow,group,gshadow} entries for root...
[*] Beginning extraction...
[*] Backing up current rootfs to rootfs_centos_6.9...
[*] Switching to new rootfs...
[*] Writing entries of root to /etc/{passwd,shadow,group,gshadow}...

C:\WSL-Distribution-Switcher>bash

C:\WSL-Distribution-Switcher>switch.py
usage: ./switch.py image[:tag]

The following distributions are currently installed:

  - centos:6.6*
  - centos:6.9
  - ubuntu:trusty

To switch back to the default distribution, specify ubuntu:trusty as the argument.

C:\WSL-Distribution-Switcher>switch.py ubuntu:trusty
[*] Probing the Linux subsystem...
[*] Moving current rootfs to rootfs_centos_6.6...
[*] Moving desired rootfs_ubuntu_trusty to rootfs...

C:\WSL-Distribution-Switcher>bash
root@NANDER:/mnt/c/WSL-Distribution-Switcher#

[Request] Add support for Kali Linux images

Title says it all :3

how do we run alpine?

get-prebuilt alpine:edge

it has no bash, how exactly do we start ?

Default user unable to sudo after installing Ubuntu devel/zesty

Apparently the default user doesn't get added to the sudo group, so if you don't set a root password so you can su and add the user to the sudo group, you're stuck with a clean system with no way to install packages.

Spurious "missing default" warning

When running switch.py I get:

You do not seem to have the default distribution installed anymore.
To reinstall it, run lxrun /uninstall and lxrun /install from the command prompt.

This happens because switch.py checks for Ubuntu trusty, but lxrun /install installs xenial nowadays.

Core dumped on fedora:25, dnf install -y sudo

Steps to reproduce:

C:/> python install.py fedora:25
C:/> lxrun /setdefaultuser root
C:/> bash
$ dnf install -y sudo

During the install a Segmentation fault (Code dumped) occurs and the RPM db is corrupted.

The image fedora:24 works as expected:

C:/> python install.py fedora:24
C:/> lxrun /setdefaultuser root
C:/> bash
$ dnf install -y sudo

Installs and finishes.

goreliu's fakefakeroot hack

Hiya, @RoliSoft I noticed you pushed an update today to support Arch, and I saw that you added @goreliu 's hack to get chroot working.

He also wrote a little hack to get fakeroot (needed for makepkg) working, where he replaces the fakeroot executable with a shellscript that quits. Would you add that too?

Also, I like how you stored the binary of the fake libchroot as text, that was a cute trick.

Also, side note, Arch also uses the wheel group for sudoer by default.

Error installing Fedora [Specified Module could not be found

PS C:\Users\knigh> C:\Windows\py.exe C:\Users\knigh\Git\WSL-Distribution-Switcher\install.py fedora
[] Probing the Linux subsystem...
[!] The Linux subsystem is currently running. Please kill all instances before continuing.
PS C:\Users\knigh> C:\Windows\py.exe C:\Users\knigh\Git\WSL-Distribution-Switcher\install.py fedora
[] Probing the Linux subsystem...
[] Default user is thinkdigital at /home/thinkdigital.
[] Reading /etc/{passwd,shadow,group,gshadow} entries for root and thinkdigital...
[] Beginning extraction...
[!] Failed to extract archive: [WinError 126] The specified module could not be found
PS C:\Users\knigh> C:\Windows\py.exe C:\Users\knigh\Git\WSL-Distribution-Switcher\install.py fedora
[] Probing the Linux subsystem...
[] Default user is thinkdigital at /home/thinkdigital.
[] Reading /etc/{passwd,shadow,group,gshadow} entries for root and thinkdigital...
[] Beginning extraction...
[!] Failed to extract archive: [WinError 126] The specified module could not be found
PS C:\Users\knigh> C:\Windows\py.exe C:\Users\knigh\Git\WSL-Distribution-Switcher\get-prebuilt.py fedora
[] Requesting authorization token...
[] Fetching manifest info for fedora:latest...
[] Downloading layer sha256:691bc14ee27487db536172a1fcdbbf956f460d1e1e1b201828e3a2bab81c5ec8...
[] Downloading layer sha256:a3ed95caeb02ffe68cdd9fd84406680ae93d633cb16422d00e8a7c22955b46d4...
[] Rootfs archive for fedora:latest saved to rootfs_fedora_latest.tar.gz.
PS C:\Users\knigh> C:\Windows\py.exe C:\Users\knigh\Git\WSL-Distribution-Switcher\install.py .\rootfs_fedora_latest.tar.
gz
[] Probing the Linux subsystem...
[] Default user is thinkdigital at /home/thinkdigital.
[] Reading /etc/{passwd,shadow,group,gshadow} entries for root and thinkdigital...
[] Beginning extraction...
[!] Failed to extract archive: [WinError 126] The specified module could not be found

Windows Info

                     ....::::       knigh@DESKTOP-0LAUJSG
             ....::::::::::::       OS: Microsoft Windows 10 Home Insider Preview 64-bit
    ....:::: ::::::::::::::::       Kernel: 10.0.16199

....:::::::::::: :::::::::::::::: Uptime: 0d 5h 35m 44s
:::::::::::::::: :::::::::::::::: Motherboard: HP 81A1
:::::::::::::::: :::::::::::::::: Shell: PowerShell 5.1.16199.1000
:::::::::::::::: :::::::::::::::: Resolution: 1280 x 720
:::::::::::::::: :::::::::::::::: Window Manager: DWM
................ ................ Font: Segoe UI
:::::::::::::::: :::::::::::::::: CPU: Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz
:::::::::::::::: :::::::::::::::: GPU: Intel(R) HD Graphics 520
:::::::::::::::: :::::::::::::::: RAM: 7121 MB / 16295 MB (44%)
'''':::::::::::: :::::::::::::::: Disk: 198GB / 233GB (85%)
'''':::: ::::::::::::::::
''''::::::::::::
''''::::

kali linux docker image returns 401 with get-prebuild.py: unauthorized

for some reason when I try to download the docker image for kali it downloads through about 6 layers, but the last layer tells me http error 401: unauthorized.

Could I perhaps use a different tool to download the image to see if the problem persists? I'm a bit of a docker newb.

EDIT: I've tried doing it with docker pull and docker save but this saves layers in a tar.gz instead of the actual rootfs.

EDIT2: Do I simply combine the layers into a single .tar.gz? How would I go about doing this? I have the docker image but it contains layers...

Installation using Cygwin's Python is broken

Hi guys.
I am trying to install debian:latest from cygwin console, but keep getting the error:

$ python3 install.py debian:latest
[*] Probing the Linux subsystem...
Traceback (most recent call last):
File "install.py", line 34, in
basedir, lxpath = probe_wsl()
File "/cygdrive/c/Users/amadeus/Desktop/tmp/WSL-Distribution-Switcher/utils.py", line 125, in probe_wsl
basedir = os.path.join(os.getenv('LocalAppData'), 'lxss')
File "/usr/lib/python3.4/posixpath.py", line 82, in join
path += b
TypeError: unsupported operand type(s) for +=: 'NoneType' and 'str'

Steps to reproduce:

python3 get-prebuilt.py debian:latest
python3 install.py debian:latest

My python version is:

$ python3 --version
Python 3.4.3

Same thing happens for python2(2.7.10), but with a slightly different error:

$ python install.py debian:latest
[*] Probing the Linux subsystem...
Traceback (most recent call last):
File "install.py", line 34, in
basedir, lxpath = probe_wsl()
File "/cygdrive/c/Users/amadeus/Desktop/tmp/WSL-Distribution-Switcher/utils.py", line 125, in probe_wsl
basedir = os.path.join(os.getenv('LocalAppData'), 'lxss')
File "/usr/lib/python2.7/posixpath.py", line 70, in join
elif path == '' or path.endswith('/'):
AttributeError: 'NoneType' object has no attribute 'endswith'

WSL is enabled, lxrun /? shows the help text.

Thanks in advance!

pritunl/archlinux still failing the post-install hook part if you don't delete the pycache

Hi there Rolisoft,

I tried doing the steps, get-prebuilt, then install, and if I didn't delete the pycache file, the build process for pacaur would fail because it would say it couldn't find awk.

Also, there's a small error when it tries to extract the apache license.txt file at the very beginning of the install, but it doesn't seem to matter.

curl, wget, and git not working on debian jessie

Debian jessie was installed by get-source.py.
After apt-get install ... , wget, curl and git are not working properly.
See below, git clone was not successful, nothing fetched from server.

root@DESKTOP-1LAII7P:/sbin# wget
Aborted (core dumped)
root@DESKTOP-1LAII7P:/sbin# curl
Aborted (core dumped)
root@DESKTOP-1LAII7P:/sbin# git clone https://github.com/RoliSoft/WSL-Distribution-Switcher.git
Cloning into 'WSL-Distribution-Switcher'...
root@DESKTOP-1LAII7P:/sbin#

Unable to start bash after installation of openSuse

Because package installation isn't working properly on Fedora, I tried installing openSuse (using the Creator's Update). Now when opening bash I get the code "Error: 0x80070040". Based on google results, it appears the issue is usually the result of manually removing the /etc/resolv.conf symlink on WSL, so the problem could be that file isn't present on the openSUSE Docker package.

Having the script in a folder with spaces doesn't work.

I have WSL downloaded on my C drive under ~\Downloads\Anything Techy\Useful Software\WSL-Distribution-Switcher, and when I go to install a rootfs:

cp: cannot stat '/mnt/c/Users/Jorge/Downloads/Anything': No such file or directory
cp: cannot stat 'techy/Useful': No such file or directory
cp: cannot stat 'Software/WSL-Distribution-Switcher/rootfs_pritunl_archlinux_latest.tar.gz': No such file or directory

Probably a simple error to fix

Got a 404 error when trying to download Debian Jessie

Using Windows 10 b1703 and the latest code from this repo:

<redacted>\WSL-Distribution-Switcher>python get-source.py debian:jessie
[*] Fetching official-images info for debian:jessie...
[*] Fetching Dockerfile from repo debuerreotype/docker-debian-artifacts/47859e2e12eadf1ff7f449d0d8b541a68289a40f/jessie...
[!] Failed to fetch Dockerfile from https://raw.githubusercontent.com/debuerreotype/docker-debian-artifacts/47859e2e12eadf1ff7f449d0d8b541a68289a40f/jessie/Dockerfile: HTTP Error 404: Not Found

No sudo after installing Debian Stretch and root password unknown

I successfully switched to Debian Stretch, but sudo does not come installed by default and I cannot obtain root access to install it.

Unable to use bash.exe after installing centos:6

Using WSL on Windows build 10.0.14393 centos:6 installs without and warnings/errors. However, when trying to run bash.exe it silently exists with a non-zero exit code.

C:\WSL-Distribution-Switcher-master>python.exe install.py centos:6
[*] Probing the Linux subsystem...
[*] Default user is root at /root.
[*] Reading /etc/{passwd,shadow,group,gshadow} entries for root...
[*] Beginning extraction...
[*] Backing up current rootfs to rootfs_ubuntu_trusty...
[*] Switching to new rootfs...
[*] Writing entries of root to /etc/{passwd,shadow,group,gshadow}...

C:\WSL-Distribution-Switcher-master>bash.exe

C:\>echo %ERRORLEVEL%
139

C:\>

Python fails on TLS connections due to missing certificates

ive yet to try others but:

python install.py alpine:latest
or python install.py voidrootfs.tar.xz

does successfully build the rootfs but when i execute 'bash' it gives me:

Error: 0x80070002

whats up?

Any way to share files between distributions outside of home?

I want to share my java distribution (currently in /opt) between distributions. Is there a recommended way to share it? The way I see it, only home directory stays outside rootfs.

Only asking, not requesting a feature.

Thank you, Great work btw!

Sample script does not fix locale or sudo for ubuntu:devel, default sudoers dangerous

Hiya,

I played around with the scripts a bit, and I noticed that the sample script does a couple of bad things.

First, instead of adding the primary user directly to the sudoers with no password, the user should be added to the group sudo with the shell command

usermod -aG sudo $defaultusername

Second, it should install sudo if the package manager is supported.

Third, for some reason with the ubuntu:devel docker image, you need to decompress with gzip the UTF8 charmap:

gzip -d /usr/share/i18n/charmaps/UTF-8.gz

before generating the locale, or else it won't work. I don't know if this happens with the ubuntu:latest docker image, but I checked a couple of times with ubuntu:devel.

CentOS can't start services (eg. SSH)

When I try to start something like SSH, I get the following error:
Failed to get D-Bus connection: Operation not permitted

On the CentOS docker page, they mention that you need to use this pre-build and then lunch it using:
docker run --privileged --name httpd -v /sys/fs/cgroup:/sys/fs/cgroup:ro -p 80:80 -d httpd

I can download the pre-build just fine, but how would you implement the docker run command from above?

New error 0x80070040 w/WSL after using Switcher, Win build 16241

Experiencing this on 10.0.16241 Build 16241 Home, Insider Fast Ring
Everything was working fine until the latest build came down for me. I previously was using Fedora 25. The first time I experienced this I did an lxinstall /uninstall /y /full - and uninstalled WSL in Windows features, then reinstalled.

C:\Users\tronik>lxrun /install
Warning: lxrun.exe is only used to configure the legacy Windows Subsystem for Linux distribution.
Distributions can be installed by visiting the Windows Store:
https://aka.ms/wslstore

-- Beta feature --
This will install Ubuntu on Windows, distributed by Canonical and licensed under its terms available here:
https://aka.ms/uowterms

Type "y" to continue: y
Downloading from the Windows Store... 100%
Extracting filesystem, this will take a few minutes...
Please create a default UNIX user account. The username does not need to match your Windows username.
For more information visit: https://aka.ms/wslusers
Enter new UNIX username: tronik
Enter new UNIX password:
Retype new UNIX password:
passwd: password updated successfully
Installation successful!
Documentation is available at:  https://aka.ms/wsldocs

I then run bash, passwd for tronik, sudo passwd root, to change passwords.

And then... get-prebuilt.py for fedora:latest and

C:\opt\WSL-Distribution-Switcher>install.py fedora:latest
[*] Probing the Linux subsystem...
[*] Default user is tronik at /home/tronik.
[*] Reading /etc/{passwd,shadow,group,gshadow} entries for root and tronik...
[*] Removing leftover rootfs-temp...
[*] Beginning extraction...
[*] Backing up current rootfs to rootfs_ubuntu_xenial...
[*] Switching to new rootfs...
[*] Writing entries of root and tronik to /etc/{passwd,shadow,group,gshadow}...

C:\opt\WSL-Distribution-Switcher>switch.py
usage: ./switch.py image[:tag]

The following distributions are currently installed:

  - fedora:latest*
  - ubuntu:xenial

You do not seem to have the default distribution installed anymore.
To reinstall it, run lxrun /uninstall and lxrun /install from the command prompt.

C:\opt\WSL-Distribution-Switcher>bash
Error: 0x80070040
Press any key to continue...

In addition, restarting or stopping/starting lxssmanager has no effect.
As well as trying to switch back to ubuntu:xenial with switch.py yields message about Linux subsystem is currently running. Please kill all instances before continuing.

rolisoft / wsl-distribution-switcher Goto Github PK

wsl-distribution-switcher's Issues

Recommend Projects

Recommend Topics

Recommend Org