-
-
Notifications
You must be signed in to change notification settings - Fork 501
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DietPi-Software | XRDP: /etc/profile is not loaded, thus $PATH is wrong #3017
Comments
@aspinks Please paste the output of: |
|
@aspinks
And if there is still nothing reported, perhaps calling the binary manually produces some meaningful output: |
Now if it helps anything, i had nextcloud installed and running, but then it would not let me login with the correct password. I deleted the mysql and nextdoor back up folders because when i tried to reinstall it told me i had no space for it. |
@aspinks Okay how much space do you actually have free:
|
Not sure why this all has a strike through on it. |
You should wrap pasts from console as code (as I quick edited inside 😉). It's better readable and GitHub does not expend markdown code, e.g. Okay, more then enough free space. Are there any I/O errors from the drive? Lets start from fresh with MariaDB, whatever went wrong with that. It was installed right with Nextcloud, so there is not other software installed that uses MariaDB, right?
|
Details:
Steps to reproduce:
Expected behaviour:
Actual behaviour:
Extra details:
Additional logs:
[FAILED] Unable to continue, will now terminate. |
Although it only occurred on the external drives. Did you unplug them without running
This does not necessarily mean something bad, sometimes for some reasons on shutdown systemd does not succeed in unmounting the drives, however all data should be synced to disk before that automatically. But in case of system crash of course a fsck should be done. At least it is worth to run once on EFI and root partition, especially because of stranger error below:
There is something more deeply broken on the system. Does the file really not exist or is perhaps the
If the last command above does not produce any output, we can try to reinstall the related package. It should not depend on the binaries itself contains:
When did you install this system? Did you follow method 1 or 2 from here? |
root@DietPi: I installed about 3-4 months ago using method 2. |
@aspinks |
Looks like that purge worked this time: root@DietPi:~# apt purge mariadb-server-10.1 |
@aspinks
|
so i did all that and it installed mariadb sucesfully. I then went to install nextcloud and get this error now. Details:
Steps to reproduce:
Expected behaviour:
Actual behaviour:
Extra details:
Additional logs:
[FAILED] DietPi-Software | Unable to continue, DietPi-Software will now terminate. |
@aspinks
However lets reinstall the module and enable it manually:
|
root@DietPi: |
It has been manually removed??
Another thing:
|
Yes it was manually removed in the graphical desktop. I checked and at this moment have no webservers installed. So i went ahead and installed Lighttpd. |
Be very careful with removing files like this manually. If you want to get rid of a PHP module or something else, remove the APT package instead of removing its files manually. There is a bunch of issues that can raise, as you can see now. And these
Then please remove the Apache PHP module:
And note that with DietPi-Software you never need to install any webserver manually. If you choose a software title that requires one, e.g. Nextcloud, the webserver that was chosen in the main menu of dietpi-software will be installed automatically (Lighttpd by default). I still see:
but Did you remove the whole /etc/php/... directory? |
/etc/php/ was not removed, it still remains. the only directories i removed were the nextcloud and mysql manually. I ran the commands above to remove apache and it did not remove it.
|
that came out weird... root@DietPi: |
@aspinks
Try:
That is again wired, this binary is part of
|
root@DietPi: Removing php7.0-opcache (7.0.33-9+0 Removing php7.0-common (7.0.33-9+0 Removing php7.0-json (7.0.33-9+0 Removing php7.0-cli (7.0.33-9+0 Removing php7.0-readline (7.0.33-9+0 |
@aspinks
Did you do a reboot some time ago? Perhaps this helps, as there seem to be something broken with the shell session. Perhaps also |
So i did do a reboot this morning earlier but for good measure I just did another reboot. I reran the apt purge command and still issues, although they seem different. root@DietPi: Removing php7.0-cli (7.0.33-9+0 Removing php7.0-common (7.0.33-9+0 Removing php7.0-json (7.0.33-9+0 Removing php7.0-opcache (7.0.33-9+0 Removing php7.0-readline (7.0.33-9+0 |
@aspinks However lets focus on this particular one again, I am somehow interested where this leads to. Otherwise the next steps:
|
I emailed you my VPN acces. Sent from [email protected] |
@aspinks |
I emailed you my Vnc acces. Sent from [email protected] I use remote desktop to connect. It's a headless setup. |
Okay found one main issue when opening a terminal from desktop:
The path variable does not include Very strange, when running a subshell, the issue is the same, but when running "sudo -s" or "sudo -i" to create a subshell, the $PATH variable is as it should be. Seems to be a general issue for XRDP sessions: https://hichenwang.blogspot.com/2014/08/setup-environment-variables-for-xrdp-in.html |
Is there anything that can be done in the meantime to re-install nextcloud? |
@aspinks The issue indeed is XRDP as mentioned above. I fixed it by placing the following file:
But this is no optimal solution as this overwrites any custom additions to the $PATH variable. However those are rare and I gave the file a very low priority (01-) so it is overridden by any other script, either in Other solutions I found are about editing What looks like a good explanation: https://serverfault.com/questions/709777/xrdp-changing-path-environment-variable
|
@aspinks |
Ya my apologies, disregard. that was total operator error... Thanks so much for all of the help! |
@aspinks I am testing this behaviour of XRDP now on VM. |
Issue confirmed on VM with XRDP. @aspinks
Enter a strong password, "visual-only" password, not required, I don't even know what this is for. I then use this outdated but very simply portable VNC viewer (TightVNC 1.3): https://www.tightvnc.com/download/1.3.10/tightvnc-1.3.10_x86_viewer.zip Another issue:
|
Lol, to check if IPv6 is solved with newer version, I installed XRDP package from stretch-backports and now I see the /etc/profile | $PATH issue is solved by this:
@aspinks
IPv6 issue solved by this as well 👍 Will add pulling this from backports on DietPi by default. |
PR up and merged to fix this on fresh installs as well as next DietPi-Update: #3020 |
+ DietPi-Software | XRDP: On Stretch, install version from backports, which resolves missing /etc/profile (e.g. $PATH) load and failing install+service if IPv6 is disabled: #3017 + DietPi-Pre-patch | XRDP: On Stretch allow upgrade of XRDP from backports to resolve certain issues + DietPi-Pre-patch | Remove pre-patch 7 which will be applied from pre-patch_file of jessie-support branch + DietPi-Pre-patch | Simplify pre-patch 6: Every Jessie system which reaches this pre-patch_file is obviously not yet on jessie-support branch + DietPi-Pre-patch | Minor coding + CHANGELOG | Complete spell check and correction
Creating a bug report/issue
Required Information
cat /DietPi/dietpi/.version
echo $G_DISTRO_NAME
orcat /etc/debian_version
uname -a
echo $G_HW_MODEL_DESCRIPTION
or (EG: RPi3)Additional Information (if applicable)
sed -n 5p /DietPi/dietpi/.hw_model
Steps to reproduce
Expected behaviour
Actual behaviour
Details:
#1 SMP Debian 4.9.168-1+deb9u4 (2019-07-19)
Steps to reproduce:
Expected behaviour:
Actual behaviour:
Extra details:
Additional logs:
The text was updated successfully, but these errors were encountered: