wiltomap's questions

Not sure to post on the right StackExchange forum. If not, let me know!

Working environment:

• OpenSUSE Leap 15.4
• MariaDB : mariadb Ver 15.1 Distrib 10.7.7-MariaDB
• PHP 8.0.25 (cli) (built: Oct 31 2022 12:00:00) ( NTS )

Based on phpinfo(), PHP ini file is: /etc/php8/cli/php.ini.

phpMyAdmin 5.1 is working well. Installation directory is: /usr/share/phpMyAdmin (default directory created when installing via zypper install phpMyAdmin command). The "famous" option $cfg['Servers'][$i]['host'] from /etc/phpMyAdmin/config.inc.php file is set to localhost (and it does work!).

I have then upgraded phpMyAdmin version this way:

srv-bla:~ # mv /usr/share/phpMyAdmin /usr/share/phpMyAdmin.old
srv-bla:~ # mkdir /usr/share/phpMyAdmin
srv-bla:~ # wget https://files.phpmyadmin.net/phpMyAdmin/5.2.0/phpMyAdmin-5.2.0-all-languages.tar.gz
srv-bla:~ # tar -xzf phpMyAdmin-5.2.0-all-languages.tar.gz
srv-bla:~ # mv phpMyAdmin-5.2.0-all-languages/* /usr/share/phpMyAdmin/

I then restart daemons and test the connection: srv-bla:~ # systemctl restart mysqld mariadb apache2

The connection page is showing right. I then enter my credentials and I get stuck with this bloody message:

Impossible to connect to server. mysqli::real_connect(): (HY000/2002): No such file or directory

I have wandered for a while on numerous webpages. Many of them suggest to modify the config.inc.php file and to set $cfg['Servers'][$i]['host'] option to 127.0.0.1 instead of localhost. Unfortunately, this does not fix the problem for me...

Am I editing the right config.inc.php file? Actually, I can find only one on the server:

srv-bla:~ # updatedb
srv-bla:~ # locate config.inc.php
/etc/phpMyAdmin/config.inc.php
/etc/phpMyAdmin/config.inc.php.rpmnew
/etc/phpMyAdmin/config.inc.php.rpmsave

Apache logs are not friendly either. Access logs returns 200 codes only, which seems normal to me (the phpMyAdmin webpage is served properly). Error logs are empty... Mysql logs are empty also (/var/log/mysql/mysqld.log).

Or course, I have check that mysqld service (same as mariadb service) is running.

Any help or ideas would be much appreciated!

EDIT

The socket file from MariaDB point of view is:

srv-bla:~ # mariadb -u root -p
Enter password:
MariaDB [(none)]> \s
[...]
UNIX socket: /var/lib/mysql/mysql.sock
[...]

The php.ini file is configured the same way:

srv-bla:~ # cat /etc/php8/cli/php.ini | grep mysqli.default_socket
mysqli.default_socket = /var/lib/mysql/mysql.sock

To me, there is no socket issue...

I have upgraded my OpenSUSE Leap from 15.3 to 15.4.

I then want to upgrade PHP from version 7 to version 8. Before upgrade, PHP version is:

server:~ # php -v
PHP 7.4.33 (cli) (built: Nov  3 2022 12:00:00) ( NTS )
Copyright (c) The PHP Group
Zend Engine v3.4.0, Copyright (c) Zend Technologies
  with Zend OPcache v7.4.33, Copyright (c), by Zend Technologies

I first started with a classic installation:

server:~ # zypper install php8

After installation is done (no particular errors/warnings mentionned), I get this:

server:~ # php -v
-bash: /usr/bin/php: No such file or directory

I then followed the official documentation for packages management which suggests to add a specific repository. I just added option -r 1 to the install command to specify the repository ID I want to install from:

server:~ # zypper addrepo https://download.opensuse.org/repositories/home:illuusio:php-next/15.4/home:illuusio:php-next.repo
server:~ # zypper refresh
server:~ # zypper install -r 1 php8

After installation is done (no particular errors/warnings mentionned), I now get this:

server:~ # php -v
php: command not found

Both installations processes suggest to uninstall PHP7, which I did accept (I don't want to keep both php7 and php8 on my server).

I then checked these:

srv-dev01:~ # whereis php
php: /usr/share/php

srv-dev01:~ # ll /usr/share/php
total 0

I don't understand why php8 is not properly installed... I guess a link is probably needed in /usr/bin/ directory but, if so, I don't get why it is not done by the installation process... I'd like to do an install as clean as possible.

Thanks for help!

I have troubles configuring a new MySQL connection with ODBC Data Sources (64bits) utility on a Windows Server 2019. I'm using MySQL Workbench to define remote accesses as follow:

The user is granted a SELECT only access:

I have noticed that the server from which I need to initiate the connection (srv-bo) has a static IP address that had already been used in the past by another server (srv-mooc). The ancient remote server entry persisted in the domain controller DNS, so I deleted it. Since then, the ODBC Data Sources (64bits) utility keeps trying to connect with the ancient server's name (srv-mooc)... It seems that some connection informations are kept in memories somewhere but I can't figure it out.

So far, I have tried to flush the DNS cache on my Windows Server 2019 from which I initiate the connection. I have also tried to set up the connection from another user session and after server reboot. Issue is still there:

Error message:

The user credentials work fine from another Windows Server 2019, so connection details are good.

Any idea to help?

Thomas

I access to a Siemens HMI screen which natively includes a VNC Server.

I get a black screen when connecting either with TightVNC or VNC Viewer. After a few seconds, the VNC connection ends up and the viewer stops.

Surprisingly, if I connect through a VPN access from a public network (Forticlient VPN), the display is correct and I don't get any black screen at all.

I have tried to modify several parameters from the VNC client, such as encoding type, colors, display size, etc. which usually help to resolve black screen issues, but nothing helped.

Any idea on potential causes and things I could test?

Thanks!

Thomas

We're setting up RADIUS authentication for wireless network connections through a Windows Server 2012 R2 (NPS).

We have to allow both domain computers (registered in Active directory) and non-domain devices, typically Android smartphones.

Following this official documentation, the behaviour is as excepted and working fine for domain computers.

Now, we would like to set up mac address authentication for Android devices, also based on Active directory. Following several posts on this subject (like this one), we have created AD users with name and password being the mac address without colons or blank spaces (ie: bc4101d16900). We have then created another network policy within NPS configuration relative to the AD Security group containing the 'Android users'. This new policy differs from the computers policy in making reference to the 'Android users' Windows Group and not the computers Windows Group.

I'm eventually wondering if such a double authentication system is possible with a Fortigate firewall (mac-address for Android devices and computer name for domain PCs). I attach a picture showing both an overview of NPS configuration for Android devices and a smartphone screenshot when attempting to connect to the SSID.

Thanks for help or ideas!

Thomas Williamson